The incidence of clinically unrecognized myocardial infarctions among 7331 Japanese-American men in Hawaii, aged 45 to 68 years and free of coronary heart disease at entry, was studied on the basis of electrocardiographic changes between successive examinations during 6 years of follow-up. The proportion of asymptomatic myocardial infarction accounted for 33% of transmural (Q-wave) myocardial infarctions identified by temporal changes on electrocardiogram and 22% of all nonfatal infarctions ascertained by either repeated examinations or hospital surveillance. The 10-year prognosis of unrecognized infarction, in terms of mortality from all causes, cardiovascular disease, and coronary heart disease, was worse (with risk ratios of 1.5 to 1.7) than that of recognized infarction, even after adjusting for age and other possible determinants, although the differences were not statistically significant. These findings suggest that regular health check-ups with an electrocardiogram would be important to detect asymptomatic myocardial infarction and to increase the opportunity of taking secondary preventive measures. However, the conclusion should await further studies based on intervention trials to determine the comparative effects of the secondary prevention on the prognosis of clinically recognized vs unrecognized infarction.