The aim of the study was to examine the association between self-reported alcohol intake and subsequent mortality from all causes, and to examine if the effect of alcohol intake on the risk of death was modified by sex, age, body mass index, and smoking habits. In a prospective population study of 7,234 women and 6,051 men aged 30-79 years alcohol- and tobacco consumption and body mass index were assessed in the period 1976-1978, and the population was followed until 1.1.1988 for mortality. A U-shaped curve described the relation between alcohol intake and mortality. The nadir of the risk function was observed at one to six beverages per week (relative risk set at 1.00). Abstainers had a relative risk of 1.37 (95% confidence intervals: 1.20-1.56), whereas those drinking more than 70 beverages per week had a relative risk of 2.29 (1.75-3.00). Among the drinkers, the risk was significantly higher than 1 only among those drinking more than 42 beverages per week. Neither sex, age, body mass index, nor smoking significantly modified the risk function. Our findings suggest that simple messages about the benefits of total abstinence may not be appropriate.