In 1949, a swept wing version of the F-84 was created with the hope of bringing performance to the level of the F-86. The last production F-84E was fitted with a swept tail, a new wing with 38.5 degrees of leading edge sweep and 3.5 degrees of anhedral, and a J35-A-25 engine producing 5,300 pound-force (23.58 kN) of thrust. The aircraft was designated XF-96A. It flew on 3 June 1950 with Otto P. Haas at the controls. Although the airplane was capable of 602 knots (693 mph, 1,115 km/h), the performance gain over the F-84E was considered minor. Nonetheless, it was ordered into production in July 1950 as the F-84F Thunderstreak. The F-84 designation was retained because the fighter was expected to be a low-cost improvement of the straight-wing Thunderjet with over 55 percent commonality in tooling.
In the meantime, the USAF, hoping for improved high-altitude performance from a more powerful engine, arranged for the British Armstrong Siddeley Sapphire turbojet engine to be built in the United States as the Wright J65. To accommodate the larger engine, YF-84Fs with a British-built Sapphire as well as production F-84Fs with the J65 had a vertically stretched fuselage, with the air intake attaining an oval cross-section. Production delays with the F-84F forced the USAF to order a number of straight-wing F-84Gs as an interim measure.
Production quickly ran into problems. Although tooling commonality with the Thunderjet was supposed to be 55 percent, in reality only fifteen percent of tools could be reused. To make matters worse, the F-84F utilized press-forged wing spars and ribs. At the time, only three presses in the United States could manufacture these, and priority was given to the Boeing B-47 Stratojet bomber over the F-84. The YJ65-W-1 engine was considered obsolete and the improved J65-W-3 did not become available until 1954. When the first production F-84F finally flew on 22 November 1952, it differed from the service test aircraft. It had a different canopy which opened up and back instead of sliding to the rear, as well as airbrakes on the sides of the fuselage instead of the bottom of the aircraft. The aircraft was considered not ready for operational deployment due to control and stability problems. The first 275 aircraft, equipped with conventional stabilizer-elevator tailplanes, suffered from accelerated stall pitch-up and poor turning ability at combat speeds. Beginning with Block 25, the problem was ameliorated by the introduction of a hydraulically powered one-piece stabilator. A number of aircraft were also retrofitted with spoilers for improved high-speed control. As a result, the F-84F was not declared operational until 12 May 1954.
The second YF-84F prototype was completed with wing-root air intakes. These were not adopted for the fighter due to loss of thrust. However, this arrangement permitted placement of cameras in the nose and the design was adopted for the RF-84F Thunderflash reconnaissance version. The first YRF-84F was completed in February 1952. The aircraft retained an armament of four machine guns and could carry up to fifteen cameras. Innovations included computerized controls which adjusted camera settings for light, speed, and altitude, a periscope to give the pilot better visualization of the target, and a voice recorder to let the pilot narrate his observations. Being largely identical to the F-84F, the Thunderflash suffered from the same production delays and engine problems, delaying operational service until March 1954. The aircraft was retired from active duty in 1957, only to be reactivated in 1961, and finally retired from the ANG in 1972.
Several modified Thunderflashes were used in the FICON project.