We attempted to determine whether graded exercise testing (GXT) soon after a myocardial infarction (MI) using an unlimited metabolic equivalent (MET) level and target heart rate (HR) of 90% of the age-predicted maximal HR is a safe procedure and yields prognostic results different from tests that use lower level end points. We administered GXT to a random sample of 85 patients at a mean of 14.8 days post-MI. End points that included ST-segment depression (STD), angina pectoris (AP), ventricular ectopic activity, and achieved MET level during testing were evaluated as predictors of cardiac events that include death, recurrent nonfatal MI, and coronary artery bypass graft surgery. We conducted short-term follow-up monitoring for one year after the initial MI and long-term follow-up monitoring for a mean of 864.7 days. No serious complications occurred during testing. Angina pectoris, STD accompanied by AP, and achieved MET level were predictive (p less than .05) of future fatal and nonfatal cardiac events during the first year of follow-up monitoring. Fifty-nine percent of the patients with AP suffered a subsequent cardiac event within one year, whereas only 17% of those without AP experienced another cardiac event. Sixty-seven percent of the patients with both STD and AP had a subsequent cardiac event within one year, but only 22% of those without AP and STD had a subsequent cardiac event. For the long-term follow-up monitoring, the presence of AP and a low achieved MET level during testing were associated with subsequent cardiac events.(ABSTRACT TRUNCATED AT 250 WORDS)