The immediate and subsequent effects of different durations of response restriction were evaluated in a multiple schedule design. Response restriction components of 15, 30, and 60 minutes were conducted between 15 minute alone components. Levels of responding subsequent to the termination of response restriction procedures were compared to free operant levels prior to the implementation of response restriction. Responding during response restriction components reduced to near zero levels. Subsequent levels of responding were similar to or exceeded free operant baseline levels. Results are discussed in terms of potential operant mechanisms responsible for levels of responding subsequent to response restriction.