To test the hypothesis that higher levels of highdensity lipoprotein cholesterol (HDL-C) in black men than in white men may offer the former greater protection against coronary heart disease (CAD), the relation between HDL-C and 7-year incidence of CAD was examined in the 5,792 white men and in the 465 black men assigned to the usual-care group of the Multiple Risk Factor Intervention Trial. CAD events included nonfatal myocardial infarction diagnosed on the basis of serial electrocardiographic change or medical record review, and fatal CAD events including sudden CAD deaths, deaths attributed to myocardial infarction or congestive heart failure caused by CAD, and deaths associated with coronary artery bypass surgery. At baseline, mean diastolic blood pressure and prevalence of cigarette smoking were significantly higher in black men, but the reverse was true for serum cholesterol (246 vs 254 mg/dl, p < 0.01). Mean HDL-C was higher in black men than in white men (49.3 vs 41.6 mg/dl, p < 0.01), but low-density lipoprotein cholesterol (LDL-C) levels were similar (159 vs 160 mg/dl). An inverse association between HDL-C and socioeconomic status was observed in black men, whereas a direct association was observed in white men. During follow-up, small reductions occurred in HDL-C and LDL-C in both groups. No black men died of stroke; 16 black and 404 white men sustained CAD events (5.1 vs 10.4/1,000 personyears of risk). The black-white relative risk was 0.49 (p = 0.005). Cox regression analyses revealed an inverse association of HDL-C with CAD events, which was not significantly different in black men from white men. After adjustment for age, diastolic blood pressure, cigarettes per day, HDL-C and LDL-C, the black-white relative risk of CAD was 0.57 (0.35 to 0.95, 95% confidence interval). Although CAD incidence was lower in black men than in white men, different distributions of HDL-C may not account for this difference.