Community studies of the relationship between mortality and mental disorders remain equivocal with mixed and unclear results. Longitudinal prospective data from the Epidemiologic Catchment Area Program are examined for the relationship between psychopathology and mortality. Analyses were performed with mortality at 1-year follow-up as the dependent variable using logistic regression. Specifically, the odds of dying from psychiatric disorders were considered while statistically adjusting for sociodemographic characteristics and treatment in primary or specialty medical care sectors. Significant predictors of dying were being male, older, and from low income households. Subjects with drug abuse or dependence were 12.4 times (CI = 2.9, 53.0) as likely to die as those without drug abuse/dependence. The odds of dying for those with major depressive disorder were over 2.6 times (CI = 1.1, 6.0) those without major depression. Separate analyses for all five communities and for only those over the age of 55 years showed similar results.