BACKGROUND Oral amiodarone effectively suppresses ventricular arrhythmias; however, full activity may take days or weeks. In patients with frequent, life-threatening ventricular arrhythmias, this delay is not acceptable. Thus, in these patients, the speed and dosing accuracy of an intravenous formulation would be beneficial. The goal of this study was to demonstrate the efficacy of intravenous amiodarone in patients with refractory, recurrent hemodynamically destabilizing ventricular tachycardia or ventricular fibrillation by determining a dose response among three regimens. RESULTS A total of 342 patients were enrolled at 46 medical centers in the United States. Patients received one of three randomized, double-blind dose regimens delivering 125, 500, or 1000 mg during the first 24 hours. Supplemental infusions (150 mg) of intravenous amiodarone could be given to treat breakthrough ventricular arrhythmias. The key efficacy end points were the arrhythmia event rate, time to first arrhythmic event, and number of supplemental infusions administered. The event rate decreased with increasing doses: median values were 0.07, 0.04, and 0.02 events per hour for the 125-, 500-, and 1000-mg dose groups, respectively, representing a significant decrease from baseline event rates (P = .043), and approached significance in the overall test for trend (P = .067). There was a significant dose-related increase in the time to first event (trend test P = .025) and a significant dose-related decrease in the number of supplemental boluses per hour (trend test P = .043). Hypotension was the most common (26%) treatment-emergent adverse event during intravenous amiodarone therapy; there was no dose-response relationship. Seventy-eight percent of the patients survived to at least 48 hours. CONCLUSIONS Intravenous amiodarone is effective for the treatment of recurrent, life-threatening ventricular tachyarrhythmias.