There is suggestive evidence that depression increases risk of myocardial infarction (MI), but there are no prospective studies in which the measure of depression corresponds to clinical criteria. This study examines prospectively whether a major depressive episode increases the risk of incident MI and evaluates the role of psychotropic medication use in this relationship. The study is based on a follow-up of the Baltimore cohort of the Epidemiologic Catchment Area Study, a survey of psychiatric disorders in the general population. A history of major depressive episode, dysphoria (2 weeks of sadness), and psychotropic medication use were assessed in 1981, and self-reported MI was assessed in 1994. Sixty-four MIs were reported among 1551 respondents free of heart trouble in 1981. Compared with respondents with no history of dysphoria, the odds ratio for MI associated with a history of dysphoria was 2.07 (95% CI, 1.16 to 3.71), and the odds ratio associated with a history of major depressive episode was 4.54 (95% CI, 1.65 to 12.44), independent of coronary risk factors. In multivariate models, use of barbiturates, meprobamates, phenothiazines, and lithium was associated with an increased risk of MI, whereas use of tricyclic antidepressants and benzodiazepines was not. Among individuals with no history of dysphoria, only lithium use was significantly associated with MI. These data suggest that a history of dysphoria and a major depressive episode increase the risk of MI. The association between psychotropic medication use and MI is probably a reflection of the primary relationship between depression and MI.