"Born in 102 Days: The Origin Story of the P-51 Mustang"

By early 1940, North American Aviation occupied a comfortable but modest position in the British war effort. The company had been supplying the Royal Air Force with its NA-16 advanced trainer, known in American service as the AT-6 Texan and in British service as the Harvard, and had met every delivery expectation placed before it. It was a solid relationship built on a reliable product, and it had earned North American a seat at the table when the British Purchasing Commission, the body Sir Henry Self had been posted to New York to lead, began expanding its search for American combat aircraft to supplement what Britain was producing at home.

Kindelberger had his eye on that table for a reason that had nothing to do with fighters. He traveled to the British Purchasing Commission's offices in New York in early 1940 to make his pitch, and what he wanted to sell was the North American B-25 Mitchell, a medium bomber his company was already developing. Sir Henry Self listened to the bomber proposal and redirected the conversation entirely. The Curtiss-Wright plant was running at capacity and P-40s were in short supply. What the RAF needed was more fighters, and what North American had was factory floor space. Self asked if NAA would manufacture Curtiss P-40s under license.

The British had already been flying the P-40, which they called the Tomahawk, and they knew precisely what it could and could not do. When Tomahawks reached RAF units in Britain in early 1941, the Air Ministry deemed them unsuitable for air combat over northern Europe almost immediately. The single-stage Allison engine that powered the P-40 performed well up to around 15,000 feet and fell away badly above it, while the Luftwaffe was fighting at 25,000 to 30,000 feet over the Channel. The Spitfire and the Hurricane handled the defense of Britain. The P-40 was for somewhere else.

That somewhere else was North Africa, and the Middle East, and the China-Burma-India theater, where the air war was largely tactical, conducted at low and medium altitude, and where the P-40's ruggedness, its availability in numbers, and the Allison's remarkable reliability in desert conditions made it genuinely effective. The RAF's strategic bombing campaign in 1940 was night bombing at moderate altitudes. The American daylight precision bombing at 25,000 to 30,000 feet that would make the Allison's ceiling a crisis did not begin until 1942 and did not reach full intensity until 1943. In the spring of 1940, with Dunkirk unfolding and France on the verge of collapse, Britain needed aircraft in whatever quantities it could obtain from any source willing to sell. The P-40 was the only American land-based fighter available in meaningful numbers. Sir Henry Self was not asking North American to build the best fighter in the world. He was asking them to build more of the only fighter available, for theaters where its limitations would not be fatal.

James Howard Kindelberger had not arrived at this juncture by accident. Born on May 8, 1895, in Wheeling, West Virginia, the son of a steelworker, he had left school in the tenth grade and gone to work throwing pig iron for seven dollars a day. He educated himself through correspondence courses, eventually earning his way into Carnegie Institute of Technology, flew as a pilot instructor during the First World War, and then worked his way through the American aviation industry with a relentless intelligence that made up for every credential he lacked. He served as chief engineer at Douglas Aircraft, where he led development of the DC-1 and DC-2, and in 1934 became president and general manager of North American Aviation. By 1940 he was one of the shrewdest men in the American aerospace business, and he had come to the British Purchasing Commission's New York offices knowing there was more opportunity in the room than the British had yet imagined.

Edgar Schmued was born on December 30, 1899, in Hornbach, Germany, where his father Heinrich, an Austrian, had settled. Edgar inherited Austrian citizenship before eventually adopting American citizenship, and like Kindelberger he was largely self-taught as an engineer, driven by a mechanical intuition that formal schooling could not have improved. He had worked for General Motors in Brazil, came to the United States through the Fokker Aeroplane Company, and joined North American Aviation as a preliminary design engineer, rising eventually to Chief of Preliminary Design. By 1940 he had designed the front turret for the XB-21 bomber, created the NA-50 single-engine fighter for Peru, seven of which were purchased by the Peruvian Air Force and nicknamed the Torito, meaning Little Bull, and worked extensively on what would become the B-25 Mitchell. He was, by any measure, one of the most capable aircraft designers in the country, and he had been quietly frustrated for years by the gap between what the state of the art could theoretically achieve and what anyone had actually tried to build. He had been sketching his own vision of a modern fighter in whatever margins his day job allowed, waiting for a question that had not yet been asked.

Kindelberger returned to Inglewood from New York and walked over to where Schmued was working. It was March 1940. "Ed," he asked, "do we want to build P-40s here?" Schmued did not hesitate. "Well, Dutch," he said, "don't let us build an obsolete airplane. Let's build a new one. We can design and build a better one." That was the answer Kindelberger had been looking for, and it was not the improvised confidence of a man talking without a plan behind him. Schmued had been quietly developing concepts for exactly this kind of fighter since the summer of 1939. The preliminary design work, the aerodynamic thinking, the structural calculations that would make a 120-day prototype possible were already substantially underway. When Kindelberger asked the question, Schmued was not proposing to start something. He was proposing to finish it. He went back to the British Purchasing Commission with a counter-proposal that had no precedent in North American Aviation's history, backed by nearly a year of work that Sir Henry Self knew nothing about.

What Kindelberger proposed would change the course of the air war over Europe, rescue the strategic bombing campaign from near collapse, and produce what many historians consider the finest piston-engine fighter aircraft ever built. North American Aviation would not build P-40s. Instead, the company would design an entirely new fighter, superior in every measurable dimension, and deliver a flying prototype in 120 days. Sir Henry Self had a reasonable objection: North American Aviation had never designed a high-speed fighter. The Harvard trainer was the fastest aircraft the company had built. The commission was being asked to trust an unproven manufacturer with a critical war contract on the word of its president alone.

Sir Henry imposed a condition. Before the commission would commit, North American would purchase the complete engineering drawings and wind tunnel test data for the Curtiss P-40 from Curtiss-Wright. The British Air Ministry required that NAA have aerodynamic reference data at high subsonic Mach numbers from an existing high-speed fighter before proceeding. NAA agreed, purchasing the Curtiss data package for £56,000. Lee Atwood, North American's chief engineer, spent time from January through April 1940 at the British Purchasing Commission's New York offices making free-hand conceptual drawings of the proposed aircraft alongside British engineers, working through specifications and demonstrating that the company's designers had a genuine technical grasp of what the RAF needed. The commission accepted. A letter of intent was issued on April 10, 1940. The commission approved the resulting detailed design drawings on May 4, 1940, and formally ordered 320 aircraft on May 29, 1940. The contract required a flying prototype within 120 days. Separately, as a condition of the United States government approving the export of military aircraft to a foreign power, the US Army Air Corps required that two examples be delivered to them free of charge for evaluation. The British had no standing to refuse that condition. Without the export approval, there was no contract.

Curtiss-Wright, upon eventually seeing what North American built with those drawings on the table, accused NAA of plagiarism. It was the complaint of a company that had sold its competitor a reference and then watched that competitor produce something demonstrably superior. The Mustang shared the Curtiss P-40's engine and had been designed with the P-40's data available, but every aerodynamic surface, every structural decision, every line of the aircraft had been Schmued's own work. The two aircraft looked nothing alike. The accusation went nowhere.

What followed was one of the most concentrated bursts of engineering achievement in aviation history, though the full picture is more nuanced than the famous 102-day figure suggests. The clock on that 102 days started ticking on May 29, 1940, when the formal British order was signed. But Schmued's clock had started much earlier. NAA had been quietly developing the conceptual groundwork for this fighter since the summer of 1939, and by the time the British contract was signed much of the detail design was already substantially complete. The 102-day rollout was not achieved by conjuring an aircraft from nothing in three and a half months. It was achieved because Schmued had spent the better part of a year preparing the foundation before anyone outside the company knew the project existed. Kindelberger put Schmued in charge of the design, with chief engineer Lee Atwood supporting the program, and 78,000 man-hours went into building the prototype over the weeks that followed. The team worked from a design philosophy that was radical in its attention to drag reduction. Where the P-40 had been developed from an airframe originally designed around an air-cooled radial engine, giving it an unavoidably large frontal area,

The engine that would power this new fighter was not chosen because it was the best available. It was chosen because it was the only available. North American designed the airframe around the Allison V-1710 liquid-cooled V-12 because in the spring of 1940 no other practical option existed for an American manufacturer on an American timeline. The Rolls-Royce Merlin, which would eventually transform the Mustang into the war-winning weapon the world remembers, was a British engine manufactured in a British factory in a Britain that was fighting for its survival, and every Merlin Rolls-Royce could produce was already spoken for by Spitfires, Hurricanes, and Lancaster bombers. The Packard Motor Car Company's license agreement to manufacture the Merlin in the United States was not signed until September 1940, four months after the NA-73X contract, and the first Packard-built Merlin was not delivered until August 1941, a full year after the prototype's first flight. More critically, the specific engine that would ultimately save the Mustang, the Merlin 61 series with its two-stage two-speed supercharger providing high-altitude performance no Allied fighter could match, was not developed until 1941 and 1942. Schmued was designing the NA-73X at almost exactly the same moment Rolls-Royce engineers were developing the supercharger technology that would eventually rescue his creation. Neither team knew the other's work would matter so much. A contract had been offered to Ford of America in 1940 to manufacture Merlins in the United States, which would have accelerated the entire engine supply picture considerably, but Henry Ford personally cancelled the agreement, unwilling to sell materials to any foreign power engaged in a conflict. The engine supply chain for the Merlin in America simply did not exist when the Mustang was being designed. Using a Pratt & Whitney or Wright radial engine was equally impossible as an alternative, because the entire aerodynamic concept, the slim nose cowling, the laminar flow wing, the radiator thrust duct, the low-drag philosophy that made the Mustang what it was, depended entirely on the narrow frontal profile of a liquid-cooled V-12. A radial engine would have produced a better-looking P-36 and nothing more. The Allison was not a compromise. It was the only engine on the table, and the convergence of Schmued's airframe with the Merlin 61 that nobody yet knew existed was not a plan. It was an accident of timing that changed the war.

North American designed the new fighter around the slim profile of the Allison V-1710 liquid-cooled engine and a notional pilot five feet ten inches tall and weighing 140 pounds. Every surface was flush-riveted. The radiator, which in most liquid-cooled fighters was a source of significant aerodynamic drag, was positioned beneath the fuselage in a carefully shaped duct that managed airflow to minimize resistance. It was later discovered in wind tunnel testing that the hot air exiting the radiator duct actually produced a small amount of thrust, as the air accelerated through the duct and exited at greater velocity than it entered, partially offsetting the drag the installation created.

The wing was built around a new airfoil profile developed by the National Advisory Committee for Aeronautics, designed around laminar-flow principles that promised to reduce drag by as much as 25 to 50 percent compared to conventional airfoils. In theory, smooth laminar airflow would remain attached over a greater portion of the wing surface before transitioning to turbulence, dramatically reducing skin friction drag and giving the new fighter exceptional speed and range from the same engine that powered the P-40. In practice, true laminar flow proved nearly impossible to maintain under operational conditions, where rivet heads, panel seams, paint variations, rain, and insect residue could trigger early turbulence across the wing surface. What the NACA airfoil delivered in the real world was not perfect laminar flow but something nearly as valuable: a genuinely low-drag wing that gave the Mustang aerodynamic efficiency no other fighter of its generation could match. The theoretical ceiling was never reached, but the practical gains were real and substantial, and they would define the character of every variant that followed.

The landing gear was designed with a wide track of twelve feet, folding inward into the wing, giving the aircraft exceptional ground handling in crosswind conditions that plagued narrow-track contemporaries. The fuel system planned for two large self-sealing tanks in the wings, holding nearly double the fuel capacity of a Spitfire. Armament was specified as two .50-caliber machine guns mounted in the lower nose cowling to fire through the propeller arc, and four additional guns in the wings, two of each caliber, staggered to fit within the wing profile.

On September 9, 1940, 102 days after the contract was signed, the prototype, designated NA-73X and carrying civil registration NX19998, rolled out of the North American plant in Inglewood, California. It had met the 120-day deadline with eighteen days to spare. It rolled out on wheels borrowed from a North American AT-6 trainer, because the proper disk brakes had not yet arrived. It rolled out without an engine, because the Allison V-1710-F3R that NAA had been promised had been delayed at the factory, and when it finally arrived it had been modified from the original engineering drawings, requiring North American to fabricate new motor mounts before installation could proceed. The aircraft also rolled out without its armament installed, carrying only the painted outlines of the gun ports on the wing leading edges where the weapons would eventually sit.

On October 26, 1940, with the Allison finally installed and turning a three-bladed Curtiss electric propeller, Vance Breese lifted NX19998 off the runway at Mines Field in Los Angeles for its maiden flight. Breese was a freelance test pilot, brought in specifically for the early flight evaluation work. Mines Field would later be expanded and renamed Los Angeles International Airport, known today as LAX. The airplane flew beautifully. It was 25 miles per hour faster than the P-40 with the same engine. The new airframe's superiority was immediately apparent to everyone who flew it. Breese continued flying the prototype through the first four flights, accumulating careful data on the aircraft's handling and performance characteristics with each sortie.

Twenty-five days later, on the aircraft's fifth test flight, it nearly ended there.

Paul Baird Balfour was North American Aviation's Chief Test Pilot, and the fifth flight of the NA-73X was his first time at its controls. Vance Breese, who knew Balfour's habits, had been so confident about what would happen that he made a bet with North American executives that Balfour would crash the prototype on his first flight. Before the flight, Balfour declined to review the flight test procedures with Edgar Schmued, saying that one airplane was like another. During the third high-speed pass over Mines Field, the engine quit. Balfour had forgotten to switch the fuel valve to the reserve tank. He was too far from the runway to make it back and put the prototype down in a freshly plowed field west of Lincoln Boulevard, but the landing gear dug into the soft soil on rollout and the aircraft cartwheeled, coming to rest inverted. Balfour was unhurt and crawled out of the wreckage.

The Civil Aeronautics Board investigated the accident and documented the damage in dry official language: engine housing broken, both wingtips damaged, tail surfaces damaged, top of fuselage damaged, and other miscellaneous damage. The NA-73X had accumulated exactly three hours and twenty minutes of flight at the moment of impact. Vance Breese collected on his bet.

The prototype was repaired and returned to flight testing on January 11, 1941. Paul Balfour was not invited back. Robert C. Chilton was hired as the new Chief Test Pilot, and Chilton made his first flight in the NA-73X on April 3, 1941. The man who replaced Balfour would go on to fly the P-51D bubble canopy prototype on November 17, 1943, and become the test pilot most closely associated with every significant Mustang milestone that followed. The NA-73X continued in the development program until it was retired on July 15, 1941. What became of the airframe after that is not definitively documented. Most published Mustang histories record the prototype as having been scrapped, and that is the most widely accepted conclusion, though primary documentation confirming this has not been firmly established in the sources available. The Mustang's origin story had nearly ended in a furrow of California farmland because the company's own Chief Test Pilot thought he already knew everything he needed to know about an airplane he had never flown.

The production aircraft that followed, designated NA-73 for the first batch of 320 and NA-83 for a follow-on order of 300 ordered in December 1940, the second batch differing from the first primarily in replacing the round exhaust stubs of the NA-73 with broader fishtail ejector exhausts, along with a range of smaller component updates to the cowling, cockpit hood, radiator ducting, trim tabs, and hydraulic lines based on feedback from early flight testing, were delivered to the Royal Air Force as the Mustang Mark I. The Allison V-1710-39, rated at 1,150 horsepower at military power from sea level to its critical altitude of 11,800 feet, drove the production aircraft to speeds that impressed every RAF pilot who flew them. They were armed with two .50-caliber machine guns in the nose cowling and four .30-caliber guns in the wings, a mixed armament that reflected the British specification. The first Mustang Is reached RAF squadrons in October 1941, considerably behind the original delivery schedule, and the pilots who received them were immediately enthusiastic about the aircraft's handling and low-altitude performance.

The problem revealed itself as soon as pilots pushed the aircraft to altitude. The Mustang I carried a service ceiling that approached 30,000 feet, and the airframe could reach that altitude without difficulty. What it could not do was fight effectively once it arrived. Above approximately 15,000 feet, where the single-stage supercharger of the Allison engine reached the limits of its design and could no longer maintain rated manifold pressure, engine power fell away sharply. The aircraft kept flying, but the Luftwaffe fighters it might encounter at 25,000 to 30,000 feet were operating at their optimal altitudes while the Mustang's Allison was laboring well below its rated output. A Mustang pilot at high altitude was not grounded, he was simply outmatched, which in a fighter is arguably worse. The RAF assigned the Mustang I to Army Co-operation Command, tasking it with low-level tactical reconnaissance and ground attack work, roles where the Allison's limitations were irrelevant and its strengths could shine. The Mustang was doing important work, useful work, but it was not what the airframe deserved.

What the airframe was doing, it did remarkably well. The first Mustang combat mission was flown on May 10, 1942, when Flying Officer G.N. Dawson of No. 26 Squadron crossed the French coast near Berck-sur-Mer, strafed aircraft hangars, and shot up a supply train on the way home. The RAF's Army Co-operation Command quickly discovered that the Mustang I at low altitude was a genuinely dangerous aircraft, faster than any German fighter the Luftwaffe could put up at sea level, capable of outrunning anything that came after it as long as it stayed low. Tactical reports were enthusiastic. Over eighteen months of operations, RAF Allison Mustangs destroyed or severely damaged 200 locomotives and 200 barges. They accomplished this at the expense of only one Mustang shot down by an enemy fighter, five lost to ground fire, and two that simply vanished with no record of their fate. The Luftwaffe could not effectively intercept an aircraft that was faster than its fighters at zero altitude. The losses came overwhelmingly from guns pointed upward from the ground, the one thing the Mustang's speed could not outrun.

There was one persistent and somewhat embarrassing problem. The Mustang I bore a striking resemblance to the Messerschmitt Bf 109 in silhouette, close enough that RAF pilots feared their own aircraft would be shot down by Allied fighters who mistook them for the enemy. Most Mustang Is in front-line RAF service flew with bright yellow bands painted across their wings specifically to prevent friendly fire incidents. The concern was not unfounded. The resemblance was not superficial. The Mustang's angular nose, similar wing planform, and overall proportions genuinely echoed the Bf 109 closely enough that in the heat of combat, a quick look from distance could produce a fatal mistake. It was a resemblance that would cause problems well beyond the flight line, but the yellow bands were the RAF's immediate practical solution to a problem with potentially lethal consequences.

By July 1942, the first Mustang was lost in action, going down over France, almost certainly to ground fire, given that pilots were operating under standing orders not to engage enemy aircraft but to use the Mustang's superior speed to disengage and escape. The following month brought a different kind of milestone. On August 19, 1942, four British and Canadian Mustang squadrons participated in the Dieppe Raid, the ill-fated amphibious assault on the French coast that served as a costly rehearsal for the eventual D-Day landings. Pilots of No. 414 Squadron of the Royal Canadian Air Force were attacked by Focke-Wulf Fw 190s during the operation, and the Mustang flew its first true air-to-air combat. Pilot Officer Hollis H. Hills of No. 414 Squadron shot down one of the attacking Fw 190s, scoring the first aerial kill by any Mustang pilot in the entire war. Eleven Mustangs were lost during the Dieppe operation, a significant single-day toll that reflected the chaos of the raid rather than any systematic defeat of the aircraft in air-to-air combat. The Fw 190s that attacked the Mustangs had the advantage of altitude, which told the same story the altitude limitation always told: the Allison Mustang was formidable on its own terms, and could be defeated on the enemy's.

In October 1942, a RAF Mustang became the first Allied single-seat fighter based in Britain to enter German airspace since the fall of France in 1940, strafing a military camp on the Dutch border and attacking a factory and fuel storage at Lathen before returning safely. The aircraft had gone where nothing else had gone, done what it was sent to do, and come home. It was doing exactly what the airframe deserved. The engine simply was not.

The altitude limitation was not a flaw unique to the Mustang, and it was not an accident. The same Allison V-1710 in the P-39 Airacobra and the P-40 Warhawk produced identical results: excellent performance up to roughly 15,000 feet, deteriorating power above it, and an effective combat altitude well below where the European air war was being fought. The aircraft themselves could physically reach 30,000 feet and beyond, their service ceilings were not the limiting factor. The limiting factor was that an engine unable to maintain rated power above 15,000 feet sent pilots into high-altitude combat with a significant and potentially fatal performance deficit against opponents whose engines were designed for exactly those altitudes. The United States Army Air Corps had specified that the V-1710 would be a single-stage supercharged engine, on the explicit assumption that aircraft requiring high-altitude capability would use external turbosuperchargers as a second stage. The Army had, in fact, also specified at the time that high-altitude performance was not necessary in a fighter, a doctrine the air war over Europe would prove catastrophically wrong. The original XP-39 prototype had been built with a turbosupercharger fitted and showed excellent high-altitude potential. Engineers removed it to reduce drag and simplify production, and the production P-39 never flew above the Allison's ceiling again. The British rejected the P-39 for exactly this reason. The Soviets accepted it enthusiastically, because the air war on the Eastern Front was largely short-range, tactical, and conducted at low altitude, where the Allison was more than adequate.

The only American fighter program that demonstrated what properly supercharged engines could achieve at high altitude was not built around a single Allison at all. The P-38 Lightning carried two V-1710s augmented by external turbosuperchargers housed in its large twin-boom engine nacelles, which provided the second stage of compression the engine's built-in supercharger could not deliver. The twin-boom design had given Lockheed's engineers exactly the internal volume the turbocharger machinery required. No single-engine design offered anything comparable. Allison had offered the Army a two-stage mechanical supercharger option on at least two separate occasions and was told no both times. By the time the bombing campaign over Germany proved beyond any argument that 25,000 to 30,000 feet was where the war was being decided, developing a new two-stage Allison from scratch would have taken years the war did not have. The Allison's altitude limitation was not an engineering failure. It was a procurement policy decision that the realities of aerial warfare over Europe would render catastrophically obsolete. The Mustang paid the price for that decision, and the solution, when it finally came, would arrive from a direction nobody in Washington had anticipated.

Fulfilling the US government's export approval condition, North American delivered two production aircraft to the United States Army Air Corps for evaluation. These two airframes, the fourth and tenth off the production line, were designated XP-51 and carried Army serial numbers 41-038 and 41-039. Both were production aircraft built in spring and early summer 1941, and North American retained them for its own development testing before transferring them to the Army later that year. The Army had demanded these aircraft as the price of allowing the British sale to proceed and then demonstrated precisely how much it valued what it had received. When North American test pilot Bob Chilton traveled to Wright Field in October 1941 to check on the aircraft, he found the XP-51 sitting on the ramp with barely one hour of flight time logged since the day it was delivered. The Army had the best fighter in the country, had required it as a condition of export approval, and had barely bothered to fly it. Testing eventually began and the results were impressive, 382 miles per hour at 13,000 feet, clearly superior to anything the Army was currently flying. The officers at Wright Field read the test reports, noted the numbers, and ordered nothing. The Army was occupied with the P-39, the P-38, and the P-47, and the XP-51 had no champion in the building. One of those two evaluation aircraft, serial 41-038, still exists today at the EAA Museum in Oshkosh, Wisconsin, a quiet monument to one of the great missed opportunities of American military procurement. The best fighter the country had produced to that point sat on a government ramp with one hour on the clock while the world burned.

The next production variant carries a timeline detail that resolves an apparent contradiction. On July 7, 1940, a full three months before the NA-73X prototype had even flown for the first time, the United States Army placed an order for 150 aircraft designated NA-91. The order predated the Lend-Lease Act, which would not be passed until March 1941, and it was not placed because the Army had suddenly discovered enthusiasm for the Mustang as a fighter. It was placed because the complex neutrality and export laws of 1940 required American military aircraft destined for a foreign power to move through American procurement channels. The Army was acting primarily as a facilitation mechanism for British procurement, not as a genuine customer. Of the 150 NA-91 aircraft produced, 93 went to the Royal Air Force. The Army retained 57, and converted 55 of those to F-6A photo-reconnaissance aircraft fitted with cameras in the fuselage. They never intended to fly them as fighters. This is entirely consistent with the Army leaving the XP-51 evaluation aircraft sitting on a ramp at Wright Field with one hour on the clock, because to the Army in 1940 and 1941 the Mustang was someone else's airplane, useful for taking pictures, and nothing more. Two of the retained NA-91 airframes were set aside for the most consequential experiment the Mustang had yet seen.

The 150 aircraft would carry two very different names depending on which flag flew over the airfield. The United States Army Air Forces, which had come into existence in June 1941 as a reorganization of the Army Air Corps, designated these aircraft P-51 and initially applied the name Apache, apparently feeling that an American fighter ought to carry an American name. The name did not stick. The RAF had been calling the aircraft Mustang since December 1940, when the British Purchasing Commission formalized the designation, and the RAF name ultimately won out. The British called the cannon-armed variant the Mustang IA, a designation that has sometimes caused confusion with the earlier mixed-armament Mustang I. The key distinction was the armament: where the original Mustang I had carried a combination of .50-caliber nose guns and .30-caliber wing guns, the Mustang IA replaced everything with four long-barreled 20-millimeter Hispano Mk II cannons in the wings, removing the nose guns entirely and concentrating all firepower outboard.

The variant that kept the Mustang program alive in American military budgets, however, was not a fighter at all.

In April 1942, appropriations for pursuit fighters were tightly constrained, but funding for attack aircraft and dive bombers was available. Kindelberger, with the instincts of a man who had been navigating military procurement for years, recognized the opening. He proposed a straightforward modification of the P-51 airframe to produce a dedicated ground attack and dive bombing aircraft, and on April 16, 1942, the Army Air Forces placed a contract for 500 aircraft designated A-36A. This was the first order for any Mustang variant placed by the American military, and it came not because the Army had finally recognized the Mustang's potential as a fighter but because Kindelberger had found a budget line that was open.

The modifications required to turn a fighter into a dive bomber were substantial but not radical. Hydraulically operated dive brakes were installed on the upper and lower surfaces of each wing, outboard of the gun positions, designed to limit dive speeds to something the airframe and pilot could survive. Hard points were added under each wing to carry two 500-pound bombs or 75-gallon drop tanks for extended range. The belly scoop was fixed in position at the forward end rather than adjustable as on earlier variants. Armament settled on six .50-caliber machine guns, two in the nose cowling and four in the wings, all of the same caliber, eliminating the mixed armament of earlier marks. The return of nose guns on the A-36A was a deliberate choice for its ground attack mission. Where the Mustang IA had removed the nose guns entirely because a pure fighter's firepower belonged in the wings, a dive bomber and strafing aircraft benefits from guns firing through the propeller arc, giving the pilot a direct line of fire precisely aligned with the aircraft's nose during a strafing run or dive attack. The engine was the Allison V-1710-87, rated at 1,325 horsepower, the most powerful Allison yet fitted to the Mustang airframe.

Test pilots discovered during flight evaluation that the A-36A built speed in a dive with alarming enthusiasm. Initiating a vertical dive at 12,000 feet with dive brakes fully deployed, the aircraft still reached 390 miles per hour, the figure documented in Robert Gruenhagen's definitive history of the Mustang. Without brakes the speed climbed well beyond 400 miles per hour. Early pilots also discovered that extending the dive brakes immediately after peeling off caused unequal hydraulic pressure across the two sets of brakes, producing a slight roll that disrupted the pilot's aim. Proper technique, deploying the brakes in a controlled sequence, eventually cured the problem. The forces generated during pullout from a vertical dive placed severe stress on the airframe, and the permitted dive angle was ultimately restricted to 70 degrees for combat operations. The aircraft's name was as unsettled as its aerodynamics. The Army Air Forces had applied the name Apache to the A-36A, but historians have noted the difficulty in finding official documentation actually authorizing that name. The 27th Fighter-Bomber Group, one of the two primary A-36A units, circulated a petition to rename their aircraft the Invader, receiving unofficial recognition of the name. Most combat reports from both groups simply called the aircraft the Mustang. All three names appear in period documentation, and none can be definitively excluded.

Production of the A-36A was completed in March 1943, delivering all 500 aircraft. The type entered combat on June 6, 1943, when both the 27th and 86th Fighter-Bomber Groups flew missions against the Italian island of Pantelleria as part of Operation Corkscrew, an aerial and naval campaign to neutralize the garrison prior to the Allied invasion of Sicily. After Pantelleria's surrender, both groups used the island as a base for dive bombing operations against Sicily itself, beginning on July 10, 1943. The 311th Fighter-Bomber Group flew A-36As in the China-Burma-India theater from bases in Dinjan, India, attacking Japanese positions in northern Burma. In the Mediterranean theater alone, A-36As flew 23,373 combat sorties and delivered more than 8,000 tons of bombs. The type recorded 84 aerial victories, and one A-36A pilot, Lieutenant Michael T. Russo, shot down five enemy aircraft to become the only pilot known to have achieved ace status in any Allison-engined Mustang variant. RAF Allison Mustangs were deployed exclusively in Army Co-operation roles, flying tactical reconnaissance and ground attack missions designed to avoid rather than seek out air-to-air combat, making ace status essentially unattainable in that operational context regardless of aircraft capability. Russo himself was not flying a designated fighter either, but the A-36A's ground attack missions regularly brought its pilots into contact with enemy aircraft, and 84 aerial victories were credited to A-36A crews across the type's entire combat career. Russo was also the first pilot of any Mustang variant, Allison or Merlin, to reach ace status. The Germans, on the receiving end of the A-36A's dive attacks, called it the Screaming Bird, a name the aircraft's pilots considered a compliment.

The final Allison-powered Mustang variant to enter production, the P-51A, came from a contract placed on June 23, 1942 for 1,200 aircraft designated NA-99. It is worth noting that the A-36A's V-1710-87 engine carries a higher designation number than the P-51A's V-1710-81, which might suggest the -87 was a later development. In fact both engines were developed in parallel for their specific roles rather than sequentially. The V-1710-87 was optimized for low-altitude operations, rated at 1,325 horsepower at just 3,000 feet, which suited the A-36A's dive bombing and ground attack mission perfectly. The V-1710-81 took a different approach, trading peak low-altitude power for sustained mid-altitude performance. Its takeoff rating was 1,200 horsepower at 3,000 rpm, while its military power rating was 1,125 horsepower maintained all the way to 14,600 feet, a meaningful improvement over the V-1710-39's critical altitude of 11,800 feet. Neither engine was strictly superior to the other. They were purpose-built for different missions, and the P-51A is correctly called the final Allison Mustang variant by production timeline rather than by engine designation number. A-36A deliveries began in October 1942. P-51A deliveries began in March 1943. Armament was standardized to four .50-caliber Browning machine guns in the wings, two per side, with 350 rounds per gun for the inboard weapons and 280 rounds per gun for the outboard, eliminating the nose guns that had appeared on earlier variants. Drop tanks could be carried under the wings, extending ferry range to approximately 2,740 miles with the largest available tanks. The AAF School of Applied Tactics at Orlando, Florida, evaluated the P-51A and concluded it was the best American fighter available below 22,000 feet, with a top speed of 412 miles per hour at 10,000 feet and a service ceiling of 31,350 feet, a number that demonstrated the airframe's aerodynamic efficiency and illustrated precisely why the Allison's altitude limitations were so frustrating: the Mustang could reach 31,000 feet, it simply could not fight there. The British received 50 P-51As under Lend-Lease, designating them Mustang II. Another 35 were converted to F-6B reconnaissance aircraft. The original order of 1,200 was cut short at 310 delivered, because the answer to everything the Allison could not do was already on its way.

The story of the Allison variants has been told here in logical sequence, but the timeline of what happened next did not wait for those variants to reach their conclusion. The A-36A contract was signed on April 16, 1942. Fourteen days later, the Merlin conversion story began. The P-51A was not contracted until months after that. These developments were running in parallel, not in sequence, and the engine that would transform the Mustang into a war-winning weapon was being fitted to test aircraft at a Rolls-Royce facility in England while Kindelberger was still building dive bombers and photo reconnaissance aircraft around the Allison. It had begun, as so many important things in aviation history begin, with one man going for a flight.

Ronald W. Harker was a service liaison pilot for Rolls-Royce, based at the company's flight test facility at Hucknall, his job to fly operational aircraft on behalf of the company and identify opportunities where Rolls-Royce engines might improve what the RAF was flying. In late April 1942, Wing Commander Ian Campbell-Orde, commanding officer of the Air Fighting Development Unit at Duxford, called Harker and invited him to fly a Mustang. On April 30, 1942, Harker climbed into the cockpit of an RAF Mustang I, took it to altitude over the English countryside, and experienced what he later described as a revelation. The aircraft was 35 miles per hour faster than the Spitfire V at comparable power settings. The handling was exceptional. The wing generated almost no buffet. And the Allison engine, which was strangling everything this airframe was capable of, was a problem that Rolls-Royce was uniquely positioned to solve.

On May 1, 1942, Harker wrote the memorandum that would transform the Mustang from a capable low-altitude workhorse into the war-winning weapon the world remembers. His assessment was direct: the Mustang, with a powerful engine like the Merlin 61, could prove itself a formidable fighter at all altitudes, not merely the low and medium altitudes where the Allison ruled. He noted the aircraft's outstanding speed, its exceptional range, and the quality of its airframe, and argued that a Merlin installation could produce performance that no current Allied fighter could match.

Harker's memo contained one notable error, not of aeronautics but of biography. Influenced by the striking visual similarities between the Mustang and the Messerschmitt Bf 109, and aware that Schmued had been born in Germany, Harker identified the designer as a former Messerschmitt employee who had brought knowledge of German aircraft design methodology to North American. It was not true. Schmued had never worked for Messerschmitt. The visual resemblances between the two aircraft were coincidental, the product of two engineering teams independently arriving at similar solutions to the problem of building a fast low-drag fighter around a liquid-cooled engine. The myth would persist for decades, given inadvertent credibility by its appearance in Harker's famous memo, and it circulates to this day in popular accounts of the Mustang's development.

Harker consulted with Ray Dorey, the chief test engineer at Hucknall, and together they secured consent from Air Chief Marshal Sir Wilfrid Freeman to proceed with an experimental conversion program. Five Mustang airframes were obtained from the RAF and transferred to the Rolls-Royce facility, where engineers fitted each with a Merlin 65 two-stage intercooled supercharged engine. The conversion required significant redesign of the engine bay and cooling system. The Allison V-1710 used a downdraft carburetor, meaning its air intake sat on top of the nose cowling, on the dorsal surface immediately behind the propeller, the distinctive feature that identifies every Allison-powered Mustang in a photograph. The Packard Merlin used an updraft induction system, meaning the carburetor air drew from below. Moving from one engine to the other therefore required relocating the carburetor air intake from the top of the nose to the bottom, creating the distinctive chin intake that would visually define every Merlin Mustang from that point forward. The Merlin's two-stage supercharger also required an intercooler to cool the compressed charge before it entered the engine, adding further complexity to the nose redesign. The Merlin's substantially greater power output over the Allison also created a propeller problem. The three-bladed propeller that had served every Allison Mustang since the NA-73X could not efficiently absorb the additional power the Merlin produced. For its initial test flights, the Mustang X was fitted with a Spitfire IX Rotol propeller as a practical stopgap, a propeller from a completely different aircraft pressed into service because the conversion program needed to fly before a purpose-designed replacement could be developed. A larger specially designed propeller was eventually fitted as the program matured. On October 13, 1942, Rolls-Royce chief test pilot Captain R.T. Shepherd flew the first Merlin-powered Mustang, designated Mustang X, from Hucknall. The results confirmed everything Harker had predicted and exceeded them. At 30,000 feet, where the Allison-powered Mustang had been nearly useless, the Merlin Mustang was a revelation.

What Rolls-Royce was proving in England directly triggered what North American Aviation would prove in California. The British program came first, and the American program followed because of it. What made the American conversion possible was that a critical piece of the puzzle was already in place. In September 1940, long before anyone had thought about putting a Merlin in a Mustang, Rolls-Royce and the British government had signed a licensing agreement with the Packard Motor Car Company in Detroit to manufacture the Merlin in the United States. The first Packard-built Merlin ran in August 1941. The early versions powered the Curtiss P-40F Kittyhawk and the Avro Lancaster bomber. Packard had nothing to do with the Mustang at that point, but when the Rolls-Royce conversion program proved the concept in the summer of 1942, there was already an American manufacturer producing the engine in quantity. The infrastructure existed. The engine existed. All that was needed was the authorization to use it.

The progress of the Rolls-Royce Mustang X program, and Hitchcock's advocacy in Washington, led the USAAF to authorize a parallel American conversion in July 1942. Two P-51 airframes from the NA-91 production batch, the 93rd and 102nd off the line, were set aside and designated XP-78, a designation that was changed to XP-51B as the work progressed. NAA engineers in California began fitting Packard-built Merlin V-1650-3 engines, the American-manufactured version that was already in production for other aircraft. The two programs informed each other, with the teams exchanging data freely throughout the process in a spirit of friendly transatlantic collaboration that nonetheless carried a competitive edge. When Captain Shepherd lifted the first Mustang X off Hucknall's runway on October 13, 1942, beating the American effort by six weeks, Inglewood sent a letter to Hucknall congratulating the British team on winning the race.

NAA had not been sitting still while Rolls-Royce celebrated. The first XP-51B was, by NAA's own assessment, roughly an eighty percent conversion, a proof of concept rather than a finished design. The second XP-51B, which followed quickly, was close to a production-ready configuration. On November 30, 1942, Bob Chilton lifted the first XP-51B off the runway at Inglewood. Informed by what the Rolls-Royce program had already established about the Merlin's propeller requirements, NAA equipped the XP-51B with a purpose-designed four-bladed Hamilton Standard Hydromatic paddle-bladed propeller eleven feet two inches in diameter, a proper engineered solution rather than the borrowed Spitfire IX Rotol that had served the Mustang X as a stopgap. Flight test data confirmed everything the Rolls-Royce program had demonstrated and added numbers that made the most skeptical procurement officers take notice. Top speed at 29,800 feet: 440 miles per hour. Service ceiling raised by 10,000 feet over the Allison variants. The same airframe that had been effectively limited to combat below 15,000 feet by the Allison's single-stage supercharger, despite a service ceiling more than twice that altitude, was now fully competitive at 30,000 feet with the Merlin.

Rolls-Royce had hoped to convert 500 RAF Allison Mustangs to Merlin power, but the resources simply were not available in a Britain that was fighting the war on every front simultaneously. The Americans faced no such constraint. Even before either prototype had flown, the USAAF had placed an initial contract for 400 P-51Bs in August 1942, a remarkable act of confidence in a program that was still being proven. Once the XP-51B's performance figures came in and confirmed everything the Rolls-Royce program had predicted, Britain made a pragmatic decision, dropping the Mustang X program and ordering 1,000 P-51Bs under the designation Mustang III, choosing to wait for North American's production version rather than continue developing their own. The USAAF ultimately ordered 2,200 P-51Bs. The four-bladed Hamilton Standard would become the standard propeller for every production Merlin Mustang that followed. The war over Europe was about to change.

The conversion programs on both sides of the Atlantic had proven what the Merlin Mustang could do. Turning that proof into the production commitment that would actually win the air war over Europe required someone willing to fight Washington for it. That someone was already in London, and he had been fighting for this program since before either prototype had flown.

Thomas Hitchcock Jr. was not, by any conventional measure, the sort of person one expected to find reshaping American military procurement policy. Born on February 11, 1900, in Aiken, South Carolina, he had left school at seventeen to join the Lafayette Flying Corps during the First World War, was shot down behind German lines, and escaped on foot, walking more than a hundred miles through occupied France before crossing into Switzerland. He returned to the United States after the war and became, over the following two decades, the most celebrated polo player in America, earning a ten-goal handicap, the sport's highest rating, and holding it continuously from 1922 through 1940. He became a partner at Lehman Brothers, married well, and moved through the highest levels of American social life. F. Scott Fitzgerald, who knew him personally, used Hitchcock as a partial model for Tom Buchanan in The Great Gatsby, though the fictional character captured only the physical presence and none of the quality of the man.

When the United States entered the war after Pearl Harbor, Hitchcock was 42 years old, and the Army Air Forces considered him too old for a combat command. He was commissioned as a lieutenant colonel and assigned as assistant air attaché at the American Embassy in London, a role that suited the brass perfectly and Hitchcock not at all. He threw himself into it anyway, and in early 1942 he began flying the Mustang with RAF units, developing an informed opinion about what the aircraft was and what it could become. He had seen the Rolls-Royce test results. He understood what the Merlin installation meant for the aircraft's ceiling and range. And he began writing letters, reports, and memoranda to Army Air Forces brass and North American Aviation officials in Washington, making the case for the Merlin Mustang with the same focused intensity he had once brought to the polo field.

His most famous line, written in evident frustration at the indifference he encountered in Washington, captured the Mustang's predicament perfectly: "Sired by the English out of an American mother, the Mustang has no parent at Wright Field to appreciate and push its good points." It was a precise diagnosis of exactly what was wrong. The aircraft had no constituency in the American military establishment, no senior officer who had flown it and fought for it and understood what it could do. Hitchcock made himself that champion through sheer force of will, reporting the success of the Merlin test flights to anyone who would listen, predicting that the Merlin Mustang would be the best American fighter of 1943, and eventually helping to secure the USAAF commitment that would set the program in motion.

While the Inglewood plant built P-51Bs, North American's second production facility in Dallas, Texas was turning out an identical aircraft under a different designation. The P-51C, built from NAA model designations NA-103 and NA-111, was mechanically indistinguishable from the P-51B in every meaningful respect, powered by the same Packard-built Merlin and carrying the same four-gun armament. Both variants carried the RAF designation Mustang III. The first Dallas-built P-51C flew on August 5, 1943, three months after the first P-51B had flown at Inglewood. Together the two plants delivered 1,988 P-51Bs and 1,750 P-51Cs before production shifted to the definitive D model. Reconnaissance versions of the P-51C were designated F-6C. A number of both B and C models were fitted in the field with the British Malcolm Hood, a sliding canopy section developed by Aero Products Limited that improved lateral and rearward sightlines without requiring the structural modifications a full bubble canopy would demand. It was an improvement, not a solution, and everyone who flew with it knew the difference.

As the B and C models reached operational units in England in the autumn of 1943, two completely separate problems were creating hazards for pilots, and understanding them requires keeping them clearly apart rather than conflating them into one.

The first was a handling and stability problem created by the 85-gallon fuselage fuel tank added behind the cockpit beginning with the P-51B-5-NA production block. With approximately 510 pounds of fuel sitting well behind the aircraft's center of gravity, at roughly 6 pounds per gallon for 100-octane aviation gasoline, the CG shifted so far aft when the tank was full that the aircraft became, in the language of the actual pilot's manual, almost impossible to trim for hands-off level flight. The aircraft had quirky and unpredictable handling characteristics in tight turns and maneuvering flight. The manual was explicit: no aerobatics with more than 40 gallons remaining in the fuselage tank. Burn it down before combat. The problem was real and it caused accidents. It was a stability and handling problem whose consequences lived in maneuvering flight.

The second was compressibility, and it was what was killing pilots in high-speed dives. The P-51B's level flight top speed of 440 miles per hour was measured as True Airspeed at 29,800 feet, where thin air means the pilot's airspeed indicator reads considerably lower than the aircraft's actual speed through the air. In a steep dive, the aircraft descended rapidly into denser air and indicated airspeed built quickly. Above approximately Mach 0.75, reached at around 505 miles per hour indicated airspeed during dives at lower altitudes, airflow over the curved upper wing surface accelerated to transonic velocities locally, generating shock waves that disrupted the wing's lift distribution and caused violent buffeting and progressive loss of elevator authority. The distinction between Mach 0.75 and the official dive limit of Mach 0.80 is not a contradiction. Between those two points the effects were real and serious but a pilot could still fight the aircraft, still feel the controls, still attempt recovery. Above Mach 0.80, those options progressively disappeared. The NACA test report described effects beyond that threshold as increasingly dangerous, meaning that the placard was not where the problem started, it was where the problem became a likely fatality.

These were two distinct problems, and the solutions themselves confirm they were unrelated. The fuselage tank CG issue was addressed through operational procedures, specifically the requirement to burn the tank down before combat. The aircraft was manageable with that discipline and NAA saw no reason to redesign it. The compressibility problem was addressed by adding a bobweight to the elevator control system, a counterweight that applied nose-up stick force at high speed and high G loadings, preventing the aerodynamic reversal that was sending pilots into unrecoverable dives. That bobweight did nothing for the 510 pounds behind the pilot. Two different problems, two different fixes.

What most likely happened to the pilots who did not come back is this. A pilot already carrying the workload of an imperfectly trimmed aircraft due to a full fuselage tank found himself in a high-speed dive for reasons that had nothing to do with that tank, whether from combat maneuvering, an attack run, or evasive action. In that dive, compressibility took hold above Mach 0.75. The instinctive response to a dive is to pull back on the stick. With compressibility above Mach 0.80, pulling back made the dive steeper. The correct response required a pilot to consciously push forward, reduce back pressure, and let the aircraft decelerate below the compressibility threshold before attempting recovery. That response required overcoming every trained instinct in a matter of seconds while shock waves hammered the controls and the ground filled the windscreen. Most pilots simply did not have time to reach that analysis even under ideal circumstances. A pilot already spending mental bandwidth managing a poorly trimmed aircraft had even less time and even less capacity. The CG issue did not cause the dive and did not cause the compressibility. What it did was consume the margin that might otherwise have made survival possible.

The bobweight was the correct engineering response precisely because it removed the requirement to override human instinct. You cannot reliably train a pilot to push forward when everything is screaming pull back. You can design the aircraft so that pulling back does not make things worse. That is what North American eventually did. Hitchcock was not investigating the fuselage tank. He was investigating the dives, the ones the pilots could not recover from, and he was doing it the only way he knew how.

Hitchcock, serving by then as Deputy Chief of Staff of the 9th Air Support Command in charge of tactical research and development, refused to send younger men up to gather the data needed to solve the problem. At 43 he began flying the aircraft himself, going up repeatedly to probe the edges of the envelope and document what was happening. On April 18, 1944, near Salisbury in Wiltshire, one of those flights did not end. Hitchcock entered a dive from which he could not recover and was killed on impact. Two fixes eventually emerged from the data Hitchcock and others gathered at such cost. The fuselage tank problem was addressed through the pilot's manual and operational discipline, specifically the requirement to burn the tank below forty gallons before engaging in combat. For the compressibility problem, North American engineers added a bobweight to the elevator control system. A bobweight is exactly what it sounds like, a weighted mass attached to the elevator's mechanical push-pull rod that uses its own inertia under acceleration to apply a nose-up force on the stick. At high speed and high G loading, where the aerodynamic forces had been reversing the elevator's effectiveness and turning a pull into a push, the bobweight counteracted that reversal, ensuring that pulling back on the stick continued to raise the nose as the pilot expected regardless of airspeed. It did not eliminate compressibility. It removed the control betrayal that had been making compressibility unsurvivable. The procedures and the bobweight that saved the lives of hundreds of Mustang pilots in the months that followed were built on data gathered at the cost of the man who had done more than anyone outside of Rolls-Royce to put the Merlin in the airframe in the first place.

The 354th Fighter Group flew the first P-51B escort missions over Germany in December 1943. Luftwaffe pilots who had grown accustomed to watching American fighters turn back at the German border encountered something they had not seen before: American single-engine fighters at 30,000 feet over the Reich, with the range to go anywhere the bombers could go and the performance to fight the best the Luftwaffe could put up. The loss rates that had been gutting the Eighth Air Force through the summer and autumn of 1943, when missions like Schweinfurt cost sixty bombers in a single day, began to fall.

Poor rearward visibility was not a problem unique to the P-51B and C. It was an industry-wide shortcoming shared by virtually every razorback fighter of the era, from the P-47 and P-40 to the Soviet Yak-1, all of them hampered by the dorsal turtledeck built into the fuselage behind the cockpit that created a blind spot an enemy pilot could use to close undetected. By 1943 the teardrop bubble canopy had appeared on combat aircraft on both sides of the conflict. In January 1943 the USAAF sent Colonel Mark Bradley to England specifically to study the solution and find a way to bring it to American fighters. P-51B and C pilots flying from British airfields were looking across the ramp at Typhoon pilots who could see in every direction and comparing that to their own framed canopies and Malcolm hoods. The Malcolm hood was an improvement but everyone who flew with one knew the difference. The solution Bradley came back with would eventually be applied not just to the Mustang but to the P-47 Thunderbolt as well, which received its own bubble canopy beginning with the P-47D-25 production block, confirming it was the right answer for the entire American fighter fleet.

North American pulled a P-51B-1-NA, serial number 43-12102, off the Inglewood production line and handed it to their engineers with one instruction: cut down the razorback and make the bubble canopy fit. The challenge was not just structural. Forming a large compound curve in plexiglass that was optically distortion free was a manufacturing problem that had no established solution at the time. A pilot looking through a distorted canopy at 400 miles per hour trying to track an enemy aircraft could not afford visual errors introduced by the material itself. NAA first built a wooden model of the P-51 fitted with a bubble canopy for wind tunnel testing while the plexiglass forming technology was simultaneously being developed. The rear fuselage behind the cockpit was reduced in height, the new teardrop canopy was fitted, and on November 17, 1943, test pilot Bob Chilton lifted the modified aircraft off the runway at Inglewood for its first flight. The results required no interpretation. The new canopy gave the pilot something approaching true 360-degree visibility, eliminating the blind spot behind the aircraft that had been costing pilots their lives in combat. The thumbs up was immediate and unqualified.

The P-51D that went into production was more than a new canopy on an existing airframe. North American used the opportunity to address many of the limitations the B and C models had revealed in combat, though not all of them. The gun arrangement on the B and C had been a persistent problem: the four .50-caliber Brownings were tilted at sharp angles to fit within the thin wing section, creating a kink in the ammunition belts that caused frequent jams in combat. The P-51D's wing was redesigned with a longer root chord than the B and C, accommodating a different wheel door and uplock design and adding structural strength to handle 1,000 pounds of external load under each wing, enabling the larger 108-gallon drop tanks that gave the D model its extraordinary range. Six .50-caliber AN/M2 Browning machine guns were mounted fully upright, three per wing, with the guns positioned to feed cleanly and reliably. The inner gun carried 400 rounds and each outer gun 270 rounds, giving the D model a total of 1,880 rounds and firepower that no previous Mustang variant had approached. Provisions were added for up to ten five-inch High Velocity Aircraft Rockets under the wings, and the engine was upgraded to the Packard-built V-1650-7, a licensed version of the Rolls-Royce Merlin 66, producing 1,490 horsepower at normal power and 1,720 horsepower at war emergency power. The cockpit received the K-14 computing gyroscopic gunsight, based on a British Ferranti design, which pilots described when they first saw it as almost miraculous.

Two problems from the B and C models were not fixed in the D. The 85-gallon fuselage tank behind the pilot was retained. The AAF had been offered a smaller 50 or 55-gallon tank that would have meaningfully reduced the aft CG problem, and chose to keep the larger tank for the range it provided. The CG issue and the operational procedures that managed it, burn the fuselage tank down before combat, carried forward unchanged from the B and C. The compressibility problem was equally persistent. The P-51D used the same NACA laminar flow airfoil as the B and C, with the same airfoil profile and the same critical Mach number. The longer root chord changed the wing's planform geometry but not the aerodynamic characteristics that determined compressibility onset. Mach 0.75 remained the threshold where the problems began, and Mach 0.80 remained the dive limit. The bobweight on the elevator control system remained the engineering answer to the control reversal, and dive speed limits remained the operational answer to compressibility. The D model was a substantially better aircraft than the B and C in almost every measurable way. These two specific problems were not among the things that improved.

The USAAF had placed an order for 2,500 P-51Ds in July 1943, four months before the prototype flew, a measure of how confident the Air Forces had become in the Mustang program. Deliveries to fighter units began in March 1944. The D model introduced one new problem of its own: cutting down the rear fuselage to accommodate the bubble canopy reduced the keel area of the aircraft, diminishing directional stability in ways that pilots noticed immediately. The solution was a dorsal fin fillet, a triangular extension of the vertical stabilizer that ran forward along the fuselage spine, restoring vertical surface area and correcting the stability deficit. It was added beginning with the P-51D Block 10 production run, and earlier aircraft were retrofitted in the field. By June 6, 1944, when Allied forces crossed the Channel for Normandy, the P-51D was in quantity service over Europe. When production finally ended, 7,956 P-51Ds had been built at Inglewood alone, making it by a wide margin the most produced variant in the Mustang family.

When the D model entered production, the Dallas plant built P-51Ds alongside Inglewood from the start. California-built aircraft carried the suffix NA, for North American Inglewood, while Dallas-built aircraft carried the suffix NT, for North American Texas. A P-51D-5-NT was identical in every respect to a P-51D-5-NA. The split designation had originally been intended to be more distinct, with plans having existed to designate Dallas-built D models as P-51E, but the AAF ultimately decided to keep the D designation with only the plant suffix to differentiate them. Dallas produced approximately 1,454 P-51D-NT aircraft before the supply chain intervened.

The Dallas plant also developed a slight variation of the bubble canopy, known informally as the Dallas Hood, which was more bulged than the standard Inglewood canopy, giving pilots marginally better lateral visibility. The canopies were interchangeable between the two factories and Dallas-produced canopies frequently found their way onto Inglewood aircraft and vice versa.

When wartime demand outpaced the supply of Hamilton Standard four-blade propellers, the Dallas facility turned to Aeroproducts as an alternative supplier, fitting their aircraft with an eleven-foot diameter four-bladed Aeroproducts hollow steel propeller in place of the Hamilton Standard unit. Aircraft produced with the Aeroproducts propeller were designated P-51K rather than P-51D. The Aeroproducts blades were easily distinguished from the Hamilton Standard by their shape: where the Hamilton Standard blades carried distinctive cuffs at the root and rounded tips, the Aeroproducts blades gradually widened toward the center of their length and tapered to a rounded point at the tip with no cuffs. Dallas produced 1,337 P-51Ks across production blocks K-1 through K-15. When Hamilton Standard propeller supplies normalized, the Dallas line returned to building P-51Ds, resuming with the D-20-NT block. The RAF designated the P-51K the Mustang IVA, receiving 594 of the type. Reconnaissance versions were designated F-6K. The Aeroproducts propeller was not well regarded by crew chiefs in the field, who found balancing problems with many units, and it was common practice during the war to replace a K model's Aeroproducts propeller with a Hamilton Standard whenever one could be sourced from supply. In every meaningful performance characteristic, the P-51K was identical to the P-51D. The different designation reflected a wartime supply chain solution, not a change in the aircraft's capabilities, and a pilot who flew one without reading the placard on the instrument panel would not have known the difference.

In July 1944, the 332nd Fighter Group of the United States Army Air Forces completed its transition to P-51 Mustangs at Ramitelli Airfield on Italy's Adriatic coast, flying in support of the Fifteenth Air Force's strategic bombing campaign against targets deep in German-occupied Europe. The men of the 332nd were African American pilots, trained at a segregated facility at Tuskegee Army Air Field in Alabama at a time when the United States military officially doubted whether Black men were capable of flying combat aircraft. They had worked their way through P-40 Warhawks, P-39 Airacobras, and P-47 Thunderbolts before receiving the Mustang, and when they painted the tails of their new fighters red, the name that bomber crews gave them, the Red Tails, became permanent. Under the command of Colonel Benjamin O. Davis Jr., the 332nd flew escort missions into Czechoslovakia, Austria, Hungary, Poland, and Germany, protecting the heavy bombers of the Fifteenth Air Force against an increasingly desperate Luftwaffe. Davis named his personal aircraft "By Request," honoring the white bomber crews who had specifically asked for the Red Tails as their escort. The 332nd flew more than 15,000 sorties during the war and lost 66 men in combat and accidents, fighting two wars simultaneously, one against the Luftwaffe and one against the institutional racism that had tried to prevent them from reaching the cockpit in the first place.

The spring of 1944 also brought the first encounters with a threat unlike anything Allied pilots had faced before. The Messerschmitt Me 262 Schwalbe, the world's first operational jet fighter, began appearing over Germany in the autumn of 1944, and the numbers it put on paper were sobering. Powered by two Junkers Jumo 004 axial-flow turbojets, the Me 262 could reach approximately 540 miles per hour at altitude, more than 100 miles per hour faster than the P-51D. Its armament of four 30-millimeter MK 108 cannons could destroy a B-17 in a single pass. The Smithsonian Institution's National Air and Space Museum described it plainly: the Me 262 could engage in, or retire from, combat at will.

What it could not do was solve Germany's problems for her.

The Jumo 004's turbine blades, machined from inferior alloys because Germany's strategic reserves of nickel and chromium were exhausted by 1944, had a service life of just ten to twenty-five hours before requiring replacement. Allied bombing of Messerschmitt's assembly plants meant completed airframes sat waiting for engines. Allied attacks on German fuel production meant the jets were often grounded for lack of fuel. And in late 1943, Adolf Hitler had personally intervened to demand the Me 262 be developed primarily as a fast bomber rather than a fighter. The popular account holds that this decision alone delayed its deployment as an interceptor by several critical months, though modern historians generally view the delay as more complicated, with engine reliability problems, fuel shortages, production disruption, and training deficiencies contributing equally alongside Hitler's bomber insistence. The bomber order was real. It was not the only obstacle. When Kommando Nowotny, the first operational Me 262 unit, finally began flying combat missions in early October 1944, it was too late, in too small numbers, against an enemy that had already learned its weakness.

The weakness was elementary. For all the Me 262's speed in level flight, its Jumo engines accelerated agonizingly slowly from idle. A pilot who reduced throttle on a landing approach was largely committed to landing, because advancing the throttles again to go around risked a dangerous delay before thrust returned. At takeoff, the jet was at its most vulnerable, building speed along a runway while its engines came up to power. And once away from the airfield, the Me 262's wide turning radius meant that a P-51 pilot who could get inside its turn held a decisive advantage regardless of the speed differential.

The USAAF response was immediate and systematic. They called it Rat Catcher. P-51 formations loitered at altitude over known Me 262 airfields, using the Mustang's endurance, which no jet of the era could match, to simply wait. When jets were detected taking off or landing, Mustang pilots rolled over and dove, using altitude to close the speed gap that level flight never could. When Me 262s returned from missions with fuel running low and ammunition expended, they slowed for the approach and the Mustangs were waiting.

On October 7, 1944, First Lieutenant Urban L. Drew of the 361st Fighter Group was leading an escort mission for a formation of B-17s over central Germany when he spotted two Me 262s from Kommando Nowotny beginning their takeoff rolls on the runway at Achmer aerodrome below him. Drew rolled his P-51D inverted and dove. By the time the two jets were airborne and still building speed, Drew was behind them at more than 400 miles per hour. He shot down the first Me 262, then walked his fire up the fuselage of the second and shattered its canopy. Both jets went down before reaching fighting speed. Drew had become the first Allied pilot to shoot down two jet aircraft in a single mission. His gun camera failed that day, and the Army Air Forces denied his claim for lack of evidence. It took 39 years, the German Luftwaffe archives, and the eyewitness testimony of Hauptmann Georg-Peter Eder, who had been on the runway at Achmer when Drew dove and who confirmed seeing a yellow-nosed P-51 kill both jets, before the United States Air Force awarded Drew the Air Force Cross on May 12, 1983.

On February 25, 1945, Mustangs of the 55th Fighter Group caught an entire Staffel of Me 262s at the moment of takeoff and destroyed six jets in a single engagement. The pattern was consistent: the most technologically advanced fighter Germany ever built was most often killed on the ground or on approach, by pilots in propeller-driven aircraft who had learned that if you could not outrun the jet you could simply wait for it to come home.

Me 262 pilots ultimately claimed 542 Allied aircraft destroyed between their first deployment and the end of the war, against approximately 100 combat losses of their own, a ratio that sounds impressive until measured against the thousands of bomber sorties that continued uninterrupted over the Reich regardless of the jets. The Me 262 was the more extraordinary aircraft. The P-51D was the more effective weapon of war. It had the range, the reliability, the numbers, and the pilots, and it had the patience to wait at altitude over a German airfield while the future of aviation taxied slowly into range.

While the D and K models were reaching pilots in quantity, Edgar Schmued had already turned his attention to what the Mustang might yet become. In February 1943, months before the P-51D prototype had even flown, Schmued had traveled to England to meet with the engineers at Supermarine and Rolls-Royce, studying the design principles behind the Spitfire and exploring the possibilities of a radically lighter Mustang built around a new generation of Merlin engine. The work he brought back to Inglewood led to a series of experimental lightweight variants, the XP-51F, the XP-51G, and the XP-51J, each exploring different combinations of reduced weight, tighter cowling, and more powerful engines. The XP-51F saved more than 1,100 pounds over the D model and reached 466 miles per hour, fitting a three-bladed Aeroproducts hollow steel propeller as part of the weight reduction program. The XP-51G, fitted with the Rolls-Royce Merlin 145, pushed that to 472 miles per hour and explored the upper limits of propeller efficiency by flying with a five-bladed propeller. The XP-51J returned to the Allison engine for experimental purposes. None of the three entered production. All three were proof of what the Mustang airframe still had in it.

The production result of that work was the P-51H, the final and fastest Mustang ever built. Powered by the Packard V-1650-9, the latest evolution of the Merlin series, fitted with a water injection system and automatic boost control that together produced a maximum of 2,218 horsepower at war emergency power, the P-51H represented everything the Mustang program had been moving toward since Schmued sketched the first lines of the NA-73X in 1940. The engine cowling was redesigned even tighter around the powerplant, further reducing frontal area and drag. The fuselage was lengthened slightly and the tail fin increased in height, correcting the directional stability issues that had required the dorsal fin addition on the D model. The P-51H retained the six-gun armament and bubble canopy of the D, carrying them in an airframe that was in many ways the final refinement of everything Schmued had been trying to build. On February 3, 1945, the first P-51H flew from Inglewood. Top speed at 25,000 feet was 487 miles per hour, making it the fastest production piston-engine aircraft ever accepted by the United States Army Air Forces, a record that stands to this day since no military service has accepted a production piston-engine aircraft faster since, and the USAAF itself ceased to exist in 1947.

The Army Air Forces ordered 1,445 P-51Hs in January 1945. The war in the Pacific ended in August before the production line could deliver more than 555. Some Pacific units received P-51Hs before the Japanese surrender, but none flew in combat. When the Korean War broke out in June 1950 and the USAF needed a long-range piston-engine fighter that could operate from Japanese bases against targets on the peninsula, the aircraft chosen was not the H but the older P-51D, redesignated F-51D, which was available in far greater numbers through Air National Guard stockpiles and proved perfectly suited to the ground attack role. The P-51H served out its career with Air National Guard units around the country and was never quite given the combat test it had been built for. Of the 555 P-51Hs ever built, only five survive today.

By the spring of 1944, with P-51Ds joining the B and C models over Europe, the Mustangs were not merely escorting bombers. They were hunting the Luftwaffe, sweeping ahead of the bomber streams to clear the airspace, strafing airfields on the way home, forcing the Germans into a defensive posture they would never escape.

Hermann Göring, interrogated after Germany's surrender, was asked when he had known the war was lost. His answer, recorded in post-war testimony and cited by historian Donald Miller in his account of the Eighth Air Force, was not the dramatic declaration of a single moment that popular history has sometimes constructed from it. It was quieter than that, and more specific. "The first time your bombers came over Hanover, escorted by fighters," Göring said, "I began to be worried. When they came with fighter escorts over Berlin, I knew the jig was up."

He was describing Mustangs.

The airplane that Kindelberger had promised in 120 days, that Schmued had designed in 102, that the RAF had flown low over hedgerows because its engine could not take it higher, that the American Army had evaluated and ignored, that a Rolls-Royce test pilot had recognized in a single afternoon flight over Duxford, that a polo player turned military attaché had fought for against institutional indifference until it broke through, that the Red Tails flew with red-painted tails over the rooftops of Europe, and that a generation of American pilots flew all the way to Berlin and back, was a product of desperation, inspiration, miscalculation, stubbornness, and luck in proportions that no one planned and no one could have predicted. The P-51 Mustang did not emerge from a grand design. It emerged from a company president who said no to building someone else's airplane and a designer who had been waiting to say yes to building his own.