The authors compared bretylium and lidocaine for reducing the incidence and persistence of ventricular fibrillation following aortic cross-clamp release performed during coronary artery bypass surgery. Thirty-three adult patients scheduled for elective bypass surgery were randomly assigned in a double-blind fashion to receive a bolus of bretylium, 10 mg/kg, lidocaine, 2 mg/kg, or saline, in equal volumes prior to the release of the aortic cross-clamp. Coronary artery bypass surgery was conducted using standard cardiopulmonary bypass (CPB) procedures with systemic cooling to 24 degrees to 28 degrees C. Temperature, arterial blood gases, and electrolytes were recorded. After clamp release, the first electrical rhythm was noted. Abnormal rhythms (ventricular fibrillation) were allowed to persist for 1 to 2 minutes, and if spontaneous conversion to a supraventricular rhythm did not occur, defibrillation with internal DC countershocks was applied. Patients were compared with respect to occurrence of ventricular fibrillation, need for DC countershocks, antiarrhythmic drugs, and inotropic support. There was no significant difference among the groups with respect to age, sex, preoperative medications, past medical histories, ejection fractions, average number of bypasses, cross-clamp times, or temperatures during bypass. The incidence of ventricular fibrillation after aortic cross-clamp removal was: saline 91%, lidocaine 64% (P less than 0.01), and bretylium 36% (P less than 0.01). The number of countershocks required to defibrillate, while lower in the bretylium group, did not reach statistical significance. After cardiopulmonary bypass, cardiac output and systemic vascular resistance were comparable. Bretylium warrants further study in this setting.