Intakes of alcohol and saturated fatty acids were determined through a dietary questionnaire from 1600 men (aged 49-66 y) in the Caerphilly Prospective Heart Disease Study. Platelet aggregation induced by thrombin adenosine disphosphate (ADP), and collagen was studied in subjects who had fasted and had not recently taken drugs affecting platelets. In subjects who drank alcohol, the odds ratio of a high response to aggregation was significantly reduced (primary ADP, P less than 0.05; secondary ADP, P less than 0.001; collagen, P less than 0.02). The significance was enhanced by adjusting for smoking and by including only the subjects with a high intake of saturated fatty acids or a low intake of polyunsaturated fatty acids. By contrast, the responsiveness to thrombin was slightly increased at all levels of alcohol consumption. We therefore suggest that part of the effects of alcohol on coronary heart disease may be mediated by a dose-dependent effect on certain platelet functions, modulated by the intake of dietary fat.