The antiarrhythmic efficacy of propranolol was evaluated in 32 patients with chronic high frequency ventricular arrhythmias in a placebo-controlled protocol. After a placebo control period, propranolol was begun and the dosage increased sequentially until arrhythmia suppression was achieved, side effects appeared, or a maximum dosage of 960 mg/day was reached. Computerized analysis of ambulatory recordings was used to quantify the arrhythmias. Twenty-four patients had 70--100% arrhythmia suppression at plasma levels ranging from 12--1100 ng/ml (end of dosing interval). Eight patients in this group had frequent episodes of ventricular tachycardia that were totally suppressed at or below the dosage that produced greater than or equal to 70% suppression of ventricular ectopic depolarizations (VEDs). A biphasic dose-response curve was seen in five patients who responded with a decrease in arrhythmia frequency in the lower ranges of dosages but had increased frequency of ectopic rhythms as the dosage was increased above the optimal level. Only one-third of patients responded at doses less than or equal to 160 mg/day. However, with dosages of 200--640 mg/day, an additional 40% responded. Propranolol appears to control ventricular arrhythmias safely and effectively in many patients. The finding that the antiarrhythmic effect in many patients required plasma concentrations greater than those that produce substantial beta-adrenergic blockage raises a question whether blockade of cardiac beta receptors can directly account for all of the antiarrhythmic actions of propranolol.