The safety of sotalol was evaluated in 3,257 patients treated for cardiac arrhythmias in double-blind and open-label clinical trials that support United States registration of the drug. In this composite population, 80% of patients had structural heart disease and 42% had life-threatening ventricular arrhythmias, i.e., ventricular tachycardia (VT) or fibrillation (VF). Proarrhythmia was reported in 141 patients (4.3%). Of these, 78 (2.4%) had torsades de pointes and 26 (0.8%) had sustained VT or VF. The overall incidence was higher in patients treated for sustained VT or VF (6.5%). In these patients, serious proarrhythmia was predominantly torsades de pointes (4.1%) and was more prevalent in patients with congestive heart failure and low ejection fraction. Torsades de pointes was observed early in the course of treatment, and its occurrence was related to dose. The overall mortality in patients treated with sotalol was 4.3% (139 patients); in patients with life-threatening arrhythmias, cardiac mortality was 4.8%. In only 27 patients (0.8%) was the death thought to be potentially drug-related. The deaths were not related to dose. Data from a previously reported placebo-controlled postmyocardial infarction trial indicated no significant difference in mortality between sotalol and placebo. Heart failure was reported in 3.3% of patients and was most prevalent in those with a previous history of congestive heart failure, cardiomyopathy, or structural heart disease. The occurrence of heart failure was unrelated to dose or time on drug; in more than half of the patients, sotalol treatment was continued. On average, there was no decrease in ejection fraction.(ABSTRACT TRUNCATED AT 250 WORDS)