The relation of alcohol consumption and regular leisure time physical activity with nonfatal myocardial infarction and sudden coronary death in men and women of Auckland, New Zealand, was examined in a large population-based case-control analysis. Within each sex, alcohol drinkers had decreased relative risks of both myocardial infarction and sudden coronary death compared with nondrinkers. Similarly, physical activity was associated with decreased relative risks of myocardial infarction and sudden coronary death in both women and men, but only in those subjects who had been exercising for five or more years. After controlling for hypertension, cigarette smoking, and alcohol consumption, 43% (95 confidence interval (CI) = 26-60) of coronary events could be explained by lack of exposure to physical activity. This compares with the per cent of coronary events in the study population attributable to hypertension (22%; 95% CI = 17-27) or cigarette smoking (31%; 95% CI = 25-38). Although the estimation of the attributable risk for a continuous variable is affected by the cut-point used to define exposure, from a public health viewpoint, regular leisure time physical activity may be as important as the above-mentioned major coronary heart disease risk factors.