OBJECTIVE To describe the mortality and causes of death among subjects with alcohol use disorder in comparison with those without alcohol disorder and to study whether mental disorders increase mortality in alcoholics. METHODS Data were analysed from the database of the Lundby Study, comprising 3563 subjects followed from 1947 to 1997. METHODS A community-based sample was investigated in 1947 with follow-ups in 1957, 1972 and 1997. Best-estimate consensus diagnoses of mental disorders, including alcohol use disorder, were assessed. In the total cohort, 427 cases of alcohol use disorders were identified. Differences in mortality between subjects with alcohol use disorders and non-alcoholics were studied using Cox regression models and causes of death were compared between alcoholic subjects and other participants. Risk factors for mortality among the 348 individuals with alcohol use disorders and known age-of-onset were analysed by means of Cox regression analyses. RESULTS The hazard ratio for mortality was higher for alcoholics compared to other subjects in the cohort. A substantial proportion of the causes of death among the alcoholics was suicide N=27 (6.3%) (26 males, 1 female). In the multivariate models of risk factors in alcohol use disorders, anxiety disorders, psychotic disorders, alcohol induced psychotic disorders and dementia were risk factors for premature death. CONCLUSIONS The mortality risk for subjects with alcohol use disorder was increased, females were especially vulnerable. The risk for suicide was high among males with alcohol problems. Anxiety disorders and severity of alcohol use disorder turned out as risk factors for premature death.