The predictors and effects of Q wave regression following acute Q wave myocardial infarction were examined in 1965-1982 in 127 Japanese-American men who participated in a prospective epidemiologic study of cardiovascular disease. Of these 127 men, 53 (42%) showed total regression of Q waves, 17 (13%) showed partial regression, and 57 (45%) showed no Q wave regression following acute myocardial infarction. Age at myocardial infarction and location of myocardial infarction did not predict which men would undergo Q wave regression. Q wave status after myocardial infarction (total, partial, or no regression) did not predict survival or recurrence of myocardial infarction. This study found that a substantial proportion of acute myocardial infarction cases undergo Q wave regression, indicating that clinicians and investigators alike require additional evidence to identify people with previous myocardial infarction.