Little is known about the U-shaped relation between alcohol intake and health beyond findings related to cardiovascular disease. Medically certified sickness absence is a health indicator in which coronary heart disease is only a minor factor. To investigate the relation between alcohol intake and sickness absence, records regarding medically certified sick leaves from all causes were assessed for 4 years (1997-2000) in a cohort of 1,490 male and 4,952 female municipal employees in Finland. Hierarchical Poisson regression, adjusted for self-reported behavioral and biologic risk factors, psychosocial risk factors, and cardiovascular diseases, was used to estimate the rate ratios and their 95% confidence intervals, relating sickness absence to each level of alcohol consumption. For both men and women, a significant curvilinear trend was found between level of average weekly alcohol consumption and sickness absence. The rates of medically certified sickness absence were 1.2-fold higher (95% confidence interval: 1.1, 1.3) for never, former, and heavy drinkers compared with light drinkers. The U-shaped relation between alcohol intake and health is not likely to be explained by confounding due to psychosocial differences or inclusion of former drinkers in the nondrinkers category. Moderate alcohol consumption also may reduce health problems other than cardiovascular disease.