The relation between alcohol and nonfatal acute myocardial infarction (AMI) was examined in a case-control study of 89 male patients and 271 control subjects in Fukuoka, Japan. Patients admitted for the first AMI at 2 hospitals in Fukuoka City were aged 40 to 69 years, and control subjects were recruited based on the telephone directory of the city. Information on alcohol drinking and potential coronary risk factors was obtained by using a self-administered questionnaire, and past drinkers were separated from lifelong abstainers in the analysis. After adjustment for age, occupation, cigarette smoking, strenuous exercise, body mass index, hypertension, diabetes mellitus and parental heart disease, the risk of AMI was progressively less with increasing levels of alcohol consumption. With those who never drank as a referent, adjusted odds ratios for current drinkers consuming less than 30, 30 to 59, and greater than or equal to 60 ml/day of alcohol were 1.11 (95% confidence interval 0.51 to 2.42), 0.31 (0.11 to 0.83), and 0.13 (0.05 to 0.36), respectively. These findings add to the body of data showing that alcohol drinkers are less likely to have AMI.