Ventricular arrhythmias occurring in the coronary care unit are not good predictors of ventricular arrhythmias or death during follow-up. However, arrhythmias detected by 24-hour electrocardiographic recordings at the time of hospital discharge are predictive of mortality over the subsequent 2 years. At discharge, only about 20% of patients have significant ventricular arrhythmias, defined as frequent or repetitive ventricular premature depolarizations. Using programmed ventricular stimulation, which can detect significant ventricular arrhythmias in patients with very little ectopy in 24-hour electrocardiographic recordings, 20% of patients have ventricular tachycardia 2 to 6 weeks after acute myocardial infarction (AMI). Both diastolic left ventricular (LV) dysfunction in the coronary care unit (i.e., rales or pulmonary congestion) and systolic LV dysfunction (i.e., LV ejection fraction) during hospitalization for AMI are potent predictors of mortality. Two large prospective studies examining the relations between LV dysfunction, ventricular arrhythmias and mortality concluded that mechanical dysfunction and ventricular arrhythmias are independently related to mortality. This finding provides a rationale for treating patients with frequent or repetitive ventricular arrhythmias detected near the time of hospital discharge after AMI. However, no study has yet examined whether reducing ventricular arrhythmias with antiarrhythmic drugs after AMI also reduces mortality. Lacking an answer to this question and given the frequency of adverse effects with antiarrhythmic drugs, most physicians are conservative in the treatment of patients with ventricular arrhythmias after AMI.