Ideal prostheses are artificial limbs that permit physically impaired individuals freedom of movement and independence. Transfemoral prostheses have been drastically improved in the last two decades; however, despite advancements in technology and medicine, they are incapable of generating net power about joints to allow patients to regain their original mobility and improve their quality of life. From a biomechanics perspective, currently available devices assist patients by absorbing and dissipating energy and thus hinder them from achieving daily activities. Historically, one of the main challenges for powered medical assistive devices has been the actuation system that mimics the characteristics of the biological musculoskeletal system. A medical device must be light, small, and cosmetically appealing, but it must also generate forces and torques that are proportional to the individual’s weight. For industrial applications, many forms of actuators have been developed, such as electrical motors and hydraulic and pneumatic cylinders. However, none of these actuators have been shown to be optimal for powered transfemoral prostheses. Pneumatic artificial muscles (PAMs) are candidates for powering medical assistive devices. This study first conducts a comprehensive investigation of knee biomechanics. The failure of existing PAMs to meet the actuation requirements for powered transfemoral prostheses is then demonstrated. Based on the determined size, weight, kinetic, and kinematic requirements of knee articulation, a PAM design is proposed and evaluated. The proposed PAM is experimentally validated to meet the biomechanical requirements of a human knee during walking, stair ascent, and sit-to-stand movement.