In our study we compared the prognostic significance of clinical variables, laboratory results and different classification models of ventricular ectopic beats recorded by means of a pre-discharge 24 hour ambulatory electrocardiographic monitoring, in 210 survivors of acute myocardial infarction. In addition a full multivariate analysis of the factors affecting survival time was carried out using Cox's proportional hazards (multiple) regression model. Multivariate stepwise discriminant analysis identified hypertension, congestive heart failure assessed by Killip class, and the grading system for ventricular arrhythmias as the most important prognostic variables. When Moss grading system for ventricular arrhythmias was used, the relative risk was a superior as heavy Moss grading system (Moss 2 vs Moss 1 and Moss 3-4 vs Moss 1, relative risk = 1.2 and 3.6 total death, respectively). Furthermore, the hazard ratio of Moss grading system was higher utilizing as comparison patients without ventricular ectopic beats (relative risk = 1.7 for Moss 2 and 5.3 for Moss 3-4) than patients with ventricular ectopic beats less than one/hour (relative risk = 1.2 for Moss 2 and 3.7 for Moss 3-4). Thus, in survivors of acute myocardial infarction, a rational and useful ventricular ectopic beats categorization includes both frequency and the presence or absence of malignant characteristics.