In recent years, the results of large randomized and controlled trials of antiarrhythmic agents for primary and secondary prevention of ventricular tachycardia and ventricular fibrillation have changed perceptions of the actions of antiarrhythmic agents regarding both efficacy and risk. The premature termination of the CAST trials of primary prevention in postinfarct patients highlighted the proarrhythmic risk and inefficacy of the sodium channel blockers (class I action), encainide, flecainide, and moricizine, in patients at relatively low risk for death in the long term. The excess mortality with therapy was attributed to proarrhythmia due to facilitation of reentry, especially during acute ischemia. About the same time, European trials with amiodarone, a complex agent with antiadrenergic action and powerful action to prolong refractoriness (class III action), indicated enhanced survival after infarction with amiodarone but not with agents with class I action. Recent verbal reports of larger and placebo-controlled trials (EMIAT and CAM-IAT) confirm a significant reduction in arrhythmia mortality, possibly with a favorable trend in total mortality. While an older trial with dl-sotalol (class III and beta-blocking actions) showed a trend toward improved survival after infarction, a recent trial with d-sotalol in patients with recent infarction or remote infarction and heart failure was prematurely terminated because of excess mortality attributed to proarrhythmia (torsades de pointes), indicating the importance of beta-blocking properties of a class III agent. A secondary prevention trial (ESVEM) in patients surviving an episode of VT or VF showed significant superiority of dl-sotalol compared to an array of agents that block sodium channels with respect to both efficacy and tolerance. Occurrence rates of arrhythmias treated with drugs tested for efficacy either by suppression of inducible arrhythmias or by suppression of spontaneous ectopy were higher and equivalent for both testing methods. A secondary prevention trial of amiodarone and multiple agents that block sodium channels in survivors of cardiac arrest (CASCADE) showed a significant increased efficacy of amiodarone but poorer long-term tolerance compared with the other agents. Comparative analysis of the results of the various trials suggests that class III action coupled with antiadrenergic action is more efficacious in both primary and secondary prevention of life-threatening ventricular arrhythmias and that lethal proarrhythmias may be the predominant effect in attempts at primary prevention in low-risk populations due to class I or so-called pure class III action.