The last two decades have witnessed a worldwide effort to provide limb amputees with prostheses which better emulate the natural control of the normal limb. The general schema is the detection of biopotentials indicative of centralnervous-system intent followed by timely and high-fidelity transfer of such signals to servomechanisms which mimic the ablated articulations-hence the adoption of Norbert Wiener's term “cybernetic” as the control strategy. This article, a distillation of the remarks of the author on the occasion of the ALZA Distinguished Lecture of the Biomedical Engineering Society at the Federation of American Societies of Experimental Biology meeting in Anaheim, California, in April of 1980, outlines the historical development, current status, and future prospects of cybernetic prostheses against the background of conventional artificial hands, arms, and legs. The author augments his personal experience with single-degree- and multiple-degree-of-freedom, upper-extremity protheses control and multimodal artificial knee control at MIT with parallel efforts elsewhere in the United States and in Europe. The detection and processing of electromyographic and electroneural signals with comparisons of contemporary and future effectiveness and evolving interpretations of the role of sensory feedback in prostheses in relation to research on physiological movement control are also discussed. Brief descriptions of cybernetic approaches to orthotic devices for paralyzed limb reanimation are included. This article discusses evaluation of prostheses and the thus far formidable problems of technology transfer of sophisticated electronic electromechanical artificial limbs and acceptance by healthcare providers, both of which are essential to wide-scale availability to amputees.