SUMMARY The incidence of ventricular arrhythmias in the first year after acute myocardial infarction was studied prospectively using 24-hour Holter electrocardiogram tape monitoring. Of 108 patients who entered the trial, 74 completed the 12 months of study. Patients were initially studied at 2 weeks after infarction then at 3-monthly intervals. Ventricular premature beats were observed in 80 per cent of studies, with these arrhythmias occurring in 57 per cent of the total hours studied. 'Significant' ventricular arrhythmias defined as ventricular premature beats greater than 1 in 10 beats, multifocal ventricular premature beats, bigeminy or trigeminy, salvos (2 or 3 consecutive ventricular premature beats), R on T, or ventricular tachycardia (4 or more ventricular premature beats), occurred in 54 per cent of studies, with these arrhythmias occurring in 40 per cent of the total hours studied. The incidence of ventricular arrhytbmias remained fairly constant for each 3-month period while the percentage of total studied hours for which ventricular arrhythmias occurred was significantly higher after the entry study at 2 weeks.The incidence of ventricular arrhythmias was significantly higher in patients with inferior or posterior infarction after the entry study compared with those with anterior infarction (P < 04005). Patients who had cardiac failure in the coronary care unit or a peak serum aspartate anminotransferase > 200 IU/l had a higher long-term incidence of ventricular arrhythmias (P < 0.01).Four deaths occurred during the study. In all cases ventricular premature beats were observed on at least 1 period of study, with significant ventricular premature beats in 2 of the 3 patients who died suddenly.A control group of 21 normal subjects had a random 24-hour Holter study and were found to have a significantly lower incidence of ventricular arrhythmias when compared with thepost-infarctiongroup. Ventricular premature beats were observed in 48 per cent of cases (P < 04005) for 28 per cent of the total hours studied (P < 04025) and significant ventricular arrhythmias in 19 per cent of cases (P < 04005) for 35 per cent of the total hours studied.The standard electrocardiogram proved to be a poor indicator of the true incidence of ventricular premature beats found on 24-hour Holter monitoring, with 75 per cent of patients having ventricular arrhythmias and 43 per cent showing 'significant' ventricular arrhythmias when all standard electrocardiograms showed no ventricular premature beats.In patients with coronary artery disease 60 per significant coronary artery disease without evidence cent of the deaths are sudden (Kuller, 1966). of recent infarction (Schwartz and Walsh, 1971; The major mechanism of death is thought to be Schaffer and Cobb, 1975). The magnitude of this ventricular fibrillation associated with ischaemia problem has led to repeated attempts to define (Lown and Wolf, 1971;Schaffer and Cobb, 1975), factors that might predict those patients at greatest since most findings at necropsy show evidence of...