To define the role of verapamil in the treatment of ventricular tachycardia (VT), we studied 21 patients with chronic recurrent VT. Electrophysiologic studies were performed before and during intravenous infusion of verapamil (0.15 mg/kg followed by 0.005 mg/kg per min). On the basis of the mode of VT initiation and termination, we identified three groups of patients: (a) 11 patients had VT suggestive of reentry, as VT could be initiated with ventricular extrastimulation and terminated with overdrive ventricular pacing. Verapamil did not affect the inducibility and cycle length of VT. (b) 7 patients had VT suggestive of catecholamine-sensitive automaticity as VT could not be initiated with programmed electrical stimulation but could be provoked by isoproterenol infusion. Moreover, the VT could not be converted to a sustained sinus rhythm with overdrive ventricular pacing and it resolved only with discontinuing isoproterenol infusion. Verapamil exerted no effects on VT. (c) 3 patients had VT with electrophysiologic characteristics suggestive of triggered activity related to delayed afterdepolarizations. Characteristically, after attaining a range of cycle lengths, the sinus, atrial or ventricular paced rhythm could initiate VT without ventricular extrastimulation. The first beat of VT invariably occurred late in the cardiac cycle with a premature coupling interval 0-80 ms shorter than the preceding QRS cycle length; the premature coupling interval gradually decreased as the sinus, atrial or ventricular paced cycle length progressively shortened. Of note, verapamil completely suppressed VT inducibility in these three patients. These observations lead us to suggest that verapamil does not affect VT caused by reentry and catecholamine-sensitive automaticity but is effective in suppressing VT caused by triggered activity related to delayed afterdepolarizations in humans.