In 1984, the USAF HQ approved an ATF Statement of Operational Need. The ATF requirement was refined and clarified. It was assumed at the time that the new fighter would be compelled to operate from bomb damaged runways and would need the ability to take off and land in a distance considerably less than the full length. It was stipulated in the ATF requirement that the aircraft be able to operate from a runway distance of no more than 2,000ft, both for taking off and landing. The only way to accomplish this requirement at the time was to incorporate thrust reversal ability into the nozzles of the engines. Since this would be a major sacrifice in terms of weight and complexity, it was decided that the thrust reversal function would be operable during flight as well, to assist in rapid deceleration during dogfights. The US had never fielded a fighter with such capability before, so some research and testing was necessary to support the ATF design effort. Because the thrust reversers were envisaged to operate thruout the ATF's flight envelope and not just be a form of runway braking, they were considered to be a very high risk item technologically, which could undermine the intent to produce a reliable service aircraft.

In 1988, McDonnell Douglas and Pratt & Whitney joined forces to develop a technology demonstrator that would test the practicality of this concept and help to lower the technological risk. Pratt designed a totally new and different nozzle for the F100 engines that incorporated a practical thrust reverser system, but they went a step further: they took the opportunity to incorporate thrust vectoring as well. The nozzle therefore, consisted of 2 stages: immediately to the rear of the afterburner was a series of louvres dedicated to reversal, then at the extreme rear was a two dimensional swivelling paddle system to divert the jet efflux up or down at an angle. In contrast to the appearance of a conventional circular nozzle, this was a box-like structure with flat components. The nozzle could deflect the thrust 20° up or down. A two-seat F-15B Eagle was comandeered by McDD and modified to accept the new nozzle configuration, and canard foreplanes with fitted to the air intakes. Rather than design and produce a totally new canard, off-the shelf F-18 Hornet tailplanes were used. They were mounted with considerable dihedral to allow any vortex generation to clear the wing.

This modified aircraft took to the air on 7 September 1988 and was called the F-15B S/MTD. During tests, the aircraft achieved a 25% reduction in take-off distance compared to a normal F-15; with vectored thrust it could rotate and lift off the runway at speeds as low as 42 mph. It was able to land in just 1,650 ft (500 m) of runway, compared to 7,500 ft (2,300 m) for a normal F-15. Thrust reversal was tested in flight to produce rapid decelerations, and controlled flight at angles of attack up to 80° was achieved.

While designing the vectoring and reversing nozzle, P&W engineers found the problem of cooling the reverser panels (subject to the blast of the engine exhaust at full throttle) was more difficult than originally envisioned. The nozzle apparently included more than 5 miles of welds and included 4 inch diameter cooling pipes. They felt the ATF production nozzle would add 1,000-1,500lb to the empty weight of the aircraft. Full-scale testbed research showed considerable coolling difficulties, and a tendency for the efflux to adversly affect directional stability when the reversers were used in flight. There was also concern that thrust reversers would unduly complicate the task of making the ATF exhaust nozzles sufficiently stealthy. The conclusion from this test effort was that the advantages gained from thrust reversal compared to a conventional engine nozzle were outweighed by the weight, cost, and complexity.

In December 1988, the AF SPO abandoned thrust reversal and relaxed the ATF runway length requirement from 2,000 to 3,000ft, but this was too late to affect the configuration of the engines that would fly in the ATF PAV's. Engine design was ahead of aircraft design, and too much hardware had been built already. To make the necessary changes at this stage would have been too expensive and slowed the programme down. Lockheed's and Northrop's Dem/Val aircraft prototypes were sized to fit engines with thrust reversal, and no changes were made other than Northrop choosing to delete operative reverse thrust nacelle vents. The required changes to the engine nozzle only took effect on the EMD production configuration.

The F-15 involved in the tests was eventually transferred to NASA and modified again, this time with a more elegant thrust vectoring system based on a circular shape. It was renamed F-15 ACTIVE.