Repeated continuous ambulatory electrocardiography by Holter monitoring was performed after early mobilisation post-myocardial infarction in 100 patients, without ventricular extrasystoles on classical ECG. Frequent ventricular arrhythmias were recorded in 58 p. 100 cases, unrelated to the localisation of the infarct. The observation of ventricular arrhythmias in the Coronary Care Unit (CCU) was not related to the occurrence of VEs during Holter monitoring. On the other hand, VEs were less frequent (38 p. 100 compared to 68 p. 100) in patients taking amiodarone or betablockers on transfer from the CCU. The reduced activity of patients during Holter monitoring minimises the frequency of arrhythmias, and 10 p. 100 patients were observed to have VEs only after exercise tolerance testing. These VEs seem to be a bad prognostic factor; the 9 patients who died in this series all presented VEs on ambulatory monitoring (7) or during exercise tolerance testing (2). However the number of patients with triple vessel disease was greater in the group with VEs on ambulatory monitoring (57 p. 100 compared with 13 p. 100). It is difficult to assess the precise role of the VEs amongst the other risk factors of sudden death. This is not a randomised study, but it would appear that long=term amiodarone or betablockers therapy may influence the medium-term prognosis.