The purpose of this study was to determine whether trains of subthreshold high frequency conditioning stimuli (333 Hz, 1 ms duration, 2 ms interval) delivered to the canine ventricle inhibited the response to a premature stimulus (S2) more effectively than did a single subthreshold conditioning stimulus. It was found that trains of conditioning stimuli (mean 1.21 mA) inhibited the response to S2 152 ms beyond expiration of the ventricular effective refractory period, whereas a single conditioning stimulus inhibited S2 only 20 ms or less beyond the ventricular effective refractory period. In late diastole, trains of conditioning stimuli failed to inhibit S2 when the train of stimuli caused ventricular depolarization or the latter occurred in response to the next sinus impulse. Trains of conditioning stimuli did not induce ventricular arrhythmias. Lidocaine or autonomic blockade did not alter the response to trains of conditioning stimuli. Trains of conditioning stimuli or a single conditioning stimulus inhibited the response to S2 only when they were delivered at the same electrode site. By lengthening the ventricular effective refractory period, trains of conditioning stimuli could prevent or terminate tachycardias, but this possibility is constrained, at present, by the spatial limitations of the technique.