Heavy coffee consumption has been associated with increased coronary heart disease (CHD) risk although many studies have not observed any relation. We studied the effect of coffee consumption, assessed with a 4-d food record, on the incidence of nonfatal acute myocardial infarction or coronary death in a cohort of 1971 men who were 42 to 60 y old and free of symptomatic CHD at baseline in 1984-1989. During a mean follow-up of 14 y, 269 participants experienced an acute coronary event. After adjustment for age, smoking, exercise ischemia, diabetes, income, and serum insulin concentration, the rate ratios (95% CIs) in daily nondrinkers and light (375 mL or less), moderate (reference level), and heavy (814 mL or more) drinkers were 0.84 (0.41-1.72), 1.22 (0.90-1.64), 1.00, and 1.43 (1.06-1.94). To address time dependence of the effect, the analysis was repeated for 75 CHD events that occurred during the first 5 y; the respective rate ratios were 0.42 (0.06-3.10), 2.00 (1.16-3.44), 1.00, and 2.07 (1.17-3.65). Further adjustment for serum HDL and LDL cholesterol concentration, diastolic blood pressure, maximal oxygen uptake, and waist-hip ratio slightly increased the rate ratio for heavy coffee intake. Neither the brewing method (boiling vs. filtering) nor the serum LDL cholesterol concentration had any impact on the risk estimates for coffee intake. In conclusion, heavy coffee consumption increases the short-term risk of acute myocardial infarction or coronary death, independent of the brewing method or currently recognized risk factors for CHD.