The effect of coffee consumption on the incidence of coronary heart disease (CHD) was studied prospectively in a population sample of 6765 men aged 51-59 years at baseline during the period 1974-1977, and free of myocardial infarction (MI) prior to the screening. During a 7.1-year follow-up there were 230 non-fatal myocardial infarctions. 169 coronary deaths and 478 deaths from all causes. Among men who were smokers at baseline there was no relationship between either non-fatal MI or death from CHD, and coffee consumption. Among non-smokers, a weak but far from significant trend towards an increasing incidence of CHD in heavy consumers of coffee was observed. There was an inverse relationship between mortality from all causes and coffee consumption.