Responses of pigeons were maintained by a VI schedule of food reinforcement. Conditioned punishment was programmed by having these responses concurrently produce an originally neutral stimulus. The effectiveness of this response-contingent stimulus was maintained by infrequent and prearranged stimulus-shock pairings delivered independently of responses. This conditioned punishment procedure reduced the overall response rate as long as the procedure was in effect. The extent and durability of the reduction was a function of the intensity of the shock that was paired with the stimulus. Analysis of the reduction in the overall response rate revealed: (1) a reduction of responses occurring in the absence of the response-contingent stimulus, which was designated as a "punishing" effect, and (2) a reduction of responses during the response-contingent stimulus, which was designated as a "suppressive" effect.