Factors that may influence energy requirements for ventricular defibrillation include the duration of fibrillation and the mode of resuscitation. The present study assesses the effect of these influences on the energy needed for defibrillation. Dogs were anesthetized, and arterial blood pressure and Lead II of the ECG were continuously recorded. Ventricular fibrillation was electrically induced in each dog for a period of 1, 3, 5, or 9 min. Three resuscitation techniques were evaluated: precountershock artificial ventilation (AV) and closed-chest cardiac massage (CCCM); precountershock AV/CCCM and epinephrine, 1 mg IV; and countershock without preliminary AV/CCCM or epinephrine. Each animal was shocked with successive doses of 1, 2, 4, and 8 J/kg, ceasing when either electrical conversion occurred or after the maximum dose had been delivered. If defibrillation was unaccompanied by resumption of spontaneous circulation (systolic pressure greater than 60 mm Hg greater than 2 min), AV/CCCM was administered for 1 min. In general, the incidence of defibrillation was inversely proportional to the duration of fibrillation. Epinephrine had no significant effect on the energy dose needed for conversion. After 2 min of fibrillation, however, epinephrine became increasingly important for restoration of circulation. The technique of immediate countershock was effective for episodes of fibrillation limited to approximately 3 min. Regardless of therapy, for intervals of fibrillation of up to 6 min, Gompertz data curves indicated that a delivered energy of 4--5 J/kg is the approximate energy dose associated with the maximum achievable incidence of defibrillation within the limits of this experimental protocol.