We determined whether the exercise electrocardiogram predicted acute cardiac events during moderate or strenuous physical activity among 3617 asymptomatic, hypercholesterolemic men (age range, 35 to 59 years) who were followed up in the Coronary Primary Prevention Trial. Submaximal exercise test results were obtained at entry and at annual follow-up visits in years 2 through 7. ST-segment depression or elevation (greater than or equal to 1 mm or 10 microV-sec) was considered to be a positive test result. The circumstances that surrounded each nonfatal myocardial infarction and coronary heart disease death were determined through a record review. The cumulative incidence of activity-related acute cardiac events was 2% during a mean follow-up period of 7.4 years. The risk was increased 2.6-fold in the presence of clinically silent, exercise-induced, ST-segment changes at entry (95% confidence interval [Cl], 1.3 to 5.2) after adjustment for 11 other potential risk factors. Of 62 men who experienced an activity-related event, 11 had a positive test result at entry (sensitivity, 18%; 95% Cl, 8 to 27). The specificity of the entry exercise test was 92% (95% Cl, 91 to 93). The sensitivity and specificity were similar when the length of follow-up was restricted to 1 year after testing. For a newly positive test result on a follow-up visit, the sensitivity was 24% (95% Cl, 12 to 36), and the specificity was 85% (95% Cl, 84 to 86); for any positive test result during the study (mean number of tests per subject, 6.2), the sensitivity was 37% (95% Cl, 25 to 49), and the specificity was 79% (95% Cl, 77 to 80). Our findings suggested that the presence of clinically silent, exercise-induced, ischemic ST-segment changes on a submaximal test was associated with an increased risk of activity-related acute cardiac events. However, this test was not sensitive when used to predict the occurrence of activity-related events among asymptomatic, hypercholesterolemic men. For this reason, the utility of the submaximal exercise test to assess the safety of physical activity among asymptomatic men at risk of coronary heart disease is likely to be limited.