Refractory ventricular fibrillation (VF) is a complication of severe hypothermia. Despite mixed experimental data, some authors view bretylium as the drug of choice in hypothermic VF. Bretylium was removed from Advanced Cardiac Life Support guidelines, and, to date, efficacy of amiodarone in hypothermia is unknown. OBJECTIVE To compare defibrillation rates from hypothermic VF after drug therapy with amiodarone, bretylium, and placebo. METHODS This was a randomized, blinded, and placebo-controlled laboratory experiment. Thirty anesthetized dogs were mechanically ventilated and instrumented to monitor coronary perfusion pressure (CPP), rectal core temperature, and electrocardiogram (ECG). Animals were cooled to 22 degrees C or the onset of spontaneous VF. Ventricular fibrillation was induced as needed with a transthoracic AC current. Cardiopulmonary resuscitation (CPR) was initiated and animals were randomized (n = 10 each group) to receive amiodarone 10 mg/kg (A), bretylium 5 mg/kg (B), or placebo (P) intravenously. CPR was continued while monitoring for chemical defibrillation. Rewarming was limited to removal from the cold environment. After 10 minutes, up to three escalating defibrillatory shocks were administered. Hemodynamic monitoring continued after resuscitation. Return of spontaneous circulation (ROSC) was defined as a sustainable ECG rhythm generating a corresponding arterial pressure tracing lasting a minimum of 15 minutes. Sample size permitted 80% power to detect a 60% difference in conversion rate between groups. RESULTS CPR was adequate based on CPP > 15 mm Hg in all animals. Mean (+/-SD) CPP was 35.3 +/- 18.8 mm Hg with an overall lower trend in the amiodarone group (p = 0.06). Baseline variables were similar between groups. No instance of chemical defibrillation was noted. There was no significant difference in ROSC rates between groups. Resuscitation rates were: amiodarone = 1/10, bretylium = 4/10, and placebo = 3/10 (p = 0.45). CONCLUSIONS In this model of severe hypothermic VF, neither amiodarone nor bretylium was significantly better than placebo in improving the resuscitation rate.