The magnitude of changes in heart rate that could be induced by the operant conditioning of beat-to-beat intervals was determined. Shock avoidance methods were employed during 71 sessions conducted over 17 weeks in 3 free-moving mongrel dogs. Experimental periods during which conditioning was done alternated with baseline periods. Heart rate differences between baseline and experimental periods were used as indices of conditioning. Alterations in heart rate were compared among four types of sessions: control, random shock, decrement conditioning and increment conditioning. Conditioned heart rate decrements averaged 3% while heart rate increments averaged 10%. Thus, operant conditioning carried out over prolonged periods of time resulted in relatively minor changes in heart rate.