C ardiac autonomic control is profoundly deranged after acute myocardial infarction (AMI), with evidence of impaired vagal control and high levels of sympathetic activity. The extent of this derangement, in particular subnormal vagal activity, can be assessed by the measurement of baroreflex sensitivity and heart rate variability (HRV). There is considerable evidence to show that low levels of these markers of cardiac autonomic control are strongly and independently associated with an adverse prognosis. 1 It is currently widely believed that abnormal cardiac autonomic control after AMI is not merely a consequence of the infarct but actively and deleteriously influences the clinical course of the disease. The predictive value of HRV has been exclusively derived from a 24-hour Holter electrocardiogram, which is slow and expensive to record and analyze. 2 Isolated short-term (5-minute) recordings may be of use, but data for their predictive value are scarce and have been derived largely through the post hoc selection of suitable fragments of 24-hour recordings. 3 The aim of this study was to determine the prognostic significance of shortterm recordings of HRV soon after AMI.• • • Subjects were recruited from among patients admitted to coronary care units at 3 hospitals in the United Kingdom between 1998 and 2000. This was a substudy of a double-blind, placebo-controlled comparison of the effect of early (Յ48 hours) and late (at 5 days) angiotensin-converting enzyme (ACE) inhibition with ramipril on short-term HRV. Data presented in this report are derived from HRV recordings taken before the study drug was administered. Patients were eligible if they presented Յ48 hours of AMI in sinus rhythm. AMI was diagnosed according to criteria of the World Health Organization. 4 Patients were excluded if they had established autonomic dysfunction, had a noncardiac disease likely to influence mortality rate, had a permanent pacemaker, were already treated with ACE inhibitors, or treatment with ACE inhibitors was contraindicated. Patients who were being treated with -adrenoceptor antagonists before admission were eligible for inclusion provided that there was no change to drug dose after admission but before HRV was examined. No patients began taking -adrenoceptor antagonists after admission but before HRV could be examined. All patients were discharged on an ACE inhibitor. All patients gave their written informed consent before entry into the ramipril study, which was approved by the local clinical research ethics committees. The substudy, which did not involve further patient contact, received separate approval.Studies were performed Յ48 hours of admission, with subjects in the semi-supine position after a 2-hour fast and Ն30 minutes of rest in bed. Studies were performed on the admitting ward at the bedside.