OBJECTIVE: To assess the relation between ventricular arrhythmias after myocardial infarction and left ventricular remodelling. DESIGN: Prospective study with consecutive patients. METHODS: 97 patients with acute myocardial infarction underwent serial echocardiographic examinations (days 1, 2, 3, and 7, and after 3 weeks) to determine end diastolic volume, end systolic volume, and ejection fraction; volumes were normalised for body surface area and expressed as indices. Holter monitoring was performed on the day of the final echocardiogram. Coronary angiography was performed in 88 patients before hospital discharge. RESULTS: Complex ventricular arrhythmias (defined as Lown class 3-5) were found in 16 of 97 patients. In logistic regression models, variables predictive of complex ventricular arrhythmias were end systolic volume index on admission (b = 0.054, P = 0.015) and end diastolic volume index after three weeks (b = 0.034, P = 0.012). Complex arrhythmias were also related to the increase of end diastolic and end systolic volume indices throughout the study (F = 5.62, P = 0.046, and F = 6.42, P = 0.017, respectively by MANOVA). A two stage linear regression model of ventricular volume versus time from infarct showed that both intercept (initial volume) and slope (rate of increase) were higher for patients with complex arrhythmias in both diastole and systole (P < 0.001 for all). CONCLUSIONS: Complex ventricular arrhythmias after myocardial infarction are related to the increase of left ventricular volume rather than to depressed ejection fraction. Complex arrhythmias may be an aetiological factor linking left ventricular remodelling with higher mortality, but larger follow up studies of patients with progressive left ventricular dilatation after myocardial infarction are necessary to answer these questions.