In a multicenter study, 390 patients with sustained symptomatic ventricular tachycardia or ventricular fibrillation late after acute myocardial infarction were prospectively followed up to assess determinants of mortality and recurrence of arrhythmic events. Patients were given standard antiarrhythmic treatment, which consisted primarily of drug therapy. During a mean follow-up period of 1.9 years, 133 patients (34%) died; arrhythmic events and heart failure were the most common cause of death (41 patients [11%] died suddenly, 31 [8%] died because of recurrent ventricular tachycardia or ventricular fibrillation and 23 [6%] died of heart failure). One hundred ninety-two patients (49%) had at least one recurrent arrhythmic event; 85% of first recurrent arrhythmic events were nonfatal. Multivariate analysis of data from patients who developed the arrhythmia less than 6 weeks after infarction identified five variables as independent determinants of total mortality: 1) age greater than 70 years (risk ratio 4.5); 2) Killip class III or IV in the subacute phase of infarction (risk ratio 3.5); 3) cardiac arrest during the index arrhythmia (risk ratio 1.7); 4) anterior infarction (risk ratio 2.2); and 5) multiple previous infarctions (risk ratio 1.6). Multivariate analysis of data from patients developing the arrhythmia greater than 6 weeks after infarction identified four variables as independently predictive of total mortality: 1) Q wave infarction (risk ratio 2.1); 2) cardiac arrest during the index arrhythmia (risk ratio 1.7); 3) Killip class III or IV in the subacute phase of infarction (risk ratio 1.7); and 4) multiple previous infarctions (risk ratio 1.4). The results of the two multivariate analyses were used in a model for prediction of mortality at 1 year. The average predicted mortality rate varied considerably according to the model: for 243 patients (62%) with the lowest risk, it was 13%, corresponding to an observed mortality rate of 12%; for 92 patients (24%) with intermediate risk, it was 27%, corresponding to an observed rate of 28%; for 55 patients (14%) with the highest risk, it was 64%, corresponding to an observed rate of 54%. This study shows that patients with symptomatic ventricular tachycardia or ventricular fibrillation late after myocardial infarction who are given standard antiarrhythmic treatment have a high mortality rate. The predictive model presented identifies patients at low, intermediate and high risk of death and can be of help in designing the appropriate diagnostic and therapeutic strategy for the individual patient.