We studied whether moderate alcohol drinkers have a lower total and coronary heart disease (CHD) mortality than nondrinkers. Among 276,802 U.S. men aged 40-59 enrolled in 1959 in an American Cancer Society prospective study, 42,756 deaths, 18,771 from CHD, occurred during the following 12 years (3% of the cohort was lost to follow-up). Using nondrinkers (55.3% of the cohort) as a reference category, age- and smoking-stratified relative risks (RR) of total mortality were 0.88 for occasional drinkers, 0.84 for those drinking 1 drink per day, 0.93, 1.02, 1.08, 1.22, and 1.38 for those drinking 2, 3, 4, 5, and 6 or more drinks per day, respectively. RRs of CHD mortality were 0.86, 0.79, 0.80, 0.83, 0.74, 0.85, and 0.92, respectively. Multivariate analysis failed to identify other confounders. No changes in RRs were introduced by excluding subjects with poor health or history of chronic disease at enrollment (32.8% of the cohort) or excluding subjects who died during the first 6 years of follow-up. These data indicate an apparent protective effect of moderate alcohol intake on CHD mortality that cannot be attributed to the inclusion of subjects with CHD or related diseases into the nondrinker category.