The effects of response contingent shock on stuttering frequency were studied in three adult male stutterers. Each subject’s stuttering frequency base rate was obtained, then response contingent shock and various discriminative stimuli were introduced. The general findings were: (1) introduction of response contingent shock reduced stuttering frequency essentially to zero, while removal of shock occasioned a return to base rate frequency; (2) specific stuttering behaviors could be independently manipulated; (3) for two subjects, the shock procedure did not systematically alter word output level; (4) stuttering frequency was brought under discriminative stimulus control.