BACKGROUND. Epidemiological data on serum lipoprotein(a) (Lp(a)), a presumably strong risk factor for coronary artery disease in White populations, has mostly been derived, in Black populations, from small samples. This study compares the distribution and the determinants of serum Lp(a) in Blacks and in Whites using large representative samples and the same methods in both populations. METHODS. The distribution and the correlates of serum Lp(a) were investigated in population-based samples of 701 Blacks in the Seychelles and 634 Whites in Switzerland, aged 25-64 years. Serum Lp(a) was quantified using a commercial immunoradiometric assay. RESULTS. The distribution of serum Lp(a) was similarly skewed in both ethnic groups, but median Lp(a) concentration was about twofold higher in Blacks (210 mg/l) compared to Whites (100 mg/l). The proportions of individuals with elevated serum Lp(a) (> 300 mg/l) was about 50% higher in Blacks (37.5%) than in Whites (25.2%). In both ethnic groups, serum Lp(a) was found to correlate with total cholesterol, LDL-cholesterol and apoprotein B but not with HDL-cholesterol, alcohol intake, smoking, and body mass index. The variance in serum Lp(a) concentration explained by any combination of these factors was smaller than 5.3% in the two populations. CONCLUSIONS. The measured factors did not explain the higher levels of serum Lp(a) found in Blacks compared to Whites. These findings are consistent with the hypothesis that genetic factors account for much of the variation of serum Lp(a) in both populations.