The level walking process for an above knee (A/K) amputee with a conventional prosthesis greatly compromises the amputee’s mobility. The fact that conventional prostheses lock in hyperextension during the stance phase, in contrast to the extend-flex-extend pattern during stance for the natural limb, has been suggested as a source of the amputee’s uncosmetic gait and high energy expenditure. While the lock in hyperextension during stance provides stability to prevent buckling, it requires the person to vault over the prosthetic limb. This vaidting during level walking may cause higher vertical displacements of the body center of gravity (e.g.) and accompanying higher energy requirements for the amputee. This investigation employs an amputee-interactive prosthesis simulator system to evaluate the viability of controlling the prosthetic knee joint to follow a normal knee position pattern. In order to insure that the amputeeinteractive prosthesis simulator system does not introduce gait anomalies, the system was controlled to simulate a conventional prosthesis. This showed that the simulator system has no undesired side effects since data from walking trials with the simulator system in “conventional prosthesis mode” are very similar to data from conventional prostheses in the literature. Then, an active position control scheme which controls the prosthetic knee joint to follow a normal knee position pattern was tested by two young, active amputees in level walking trials. The subjects experienced very little difficulty in walking with the active control scheme and preferred the simulator with the active control scheme to their conventional prostheses. Measured knee power requirements for the scheme indicate that this type of control is feasible without external power sources. However, measurements of the vertical displacement of the body e.g. show little difference between gait with the active control scheme and gait with a conventional prosthesis. It appears that the increased energy requirements for A/K amputees are not due in total to the lack of the extend-Jlex-extend position profde at the prosthetic knee joint.