BACKGROUND: Programmed ventricular stimulation performed early after acute myocardial infarction allows to identify patients at risk of sudden death and sustained ventricular tachycardia with high degree of predictive accuracy. This procedure, however, because of its invasive nature, is not desirable as a screening test for large numbers of patients. Therefore, it should be performed on a smaller group of postinfarction patients preselected on the basis of noninvasive testing. The aim of the present study was to identify, early after acute myocardial infarction, any procedure among noninvasive testing, able to selected with the highest sensitivity patients at risk of sudden death and sustained ventricular tachycardia to submit to programmed ventricular stimulation. METHODS: Two hundred and sixty four consecutive patients with recent myocardial infarction were evaluated and followed during a period of 12 months. In each patient 48 epidemiological, clinical and laboratory variables were evaluated. Laboratory variables were acquired between the 7th and the 12th day after the acute event. RESULTS: Multiple linear regression analysis showed that only Killip class, the number of ventricular premature depolarizations per hour and the presence of ventricular late potentials were significantly and independently related to the occurrence of sudden death and sustained ventricular tachycardia (F = 18.7; p < 0.00001). Combinations of these variables, determined at cut off levels best discriminating two subgroups of patients at different risk of the end-point events, proved to be able to accurately predict the outcome of our patients. The presence of at least one of the following conditions: Killip class > or = 2, ventricular premature depolarizations > or = 30 per hour, ventricular late potentials allowed to identify a first subgroup of patients at risk with a sensitivity of 100% (p = 0.00007), whereas the presence, at the same time, of all the above mentioned parameters allowed to identify a second subgroup of patients at risk with a 44% of positive predictive value (p = 0.00007). CONCLUSIONS: Our findings suggest that the first subgroup of postinfarction patients selected on the basis of noninvasive testing should undergo programmed ventricular stimulation, the second might be treated by adequate antiarrhythmic therapy without undergo any further investigation, whereas the remaining patients (without late potentials, in Killip class 1 and with ventricular premature depolarizations < 30 per hour) might be discharged without any antiarrhythmic therapy.