Abstract-Plasma lipoprotein(a) [Lp(a)] levels are largely genetically determined by sequences linked to the gene encoding apolipoprotein(a) [apo(a)], the distinct protein component of Lp(a). Apo(a) is highly polymorphic in length due to variation in the numbers of a sequence encoding the apo(a) kringle 4 domain, and plasma levels of Lp(a) are inversely correlated with apo(a) size. In 2 racially homogeneous Bantu populations from Tanzania differing in their dietary habits, we found that median plasma levels of Lp(a) were 48% lower in those living on a fish diet than in those living on a vegetarian diet. Considering the relationship between apo(a) size and Lp(a) plasma concentration, we have extensively evaluated apo(a) isoform distribution in the 2 populations to determine the impact of apo(a) size in the determination of Lp(a) values. The majority of individuals (82% of the fishermen and 80% of the vegetarians) had 2 expressed apo(a) alleles. Additionally, the fishermen had a high frequency of large apo(a) isoforms, whereas a higher frequency of small isoforms was found in the vegetarians. When subjects from the 2 groups were matched for apo(a) phenotype, the median Lp(a) value was 40% lower in Bantus on the fish diet than in those on the vegetarian diet. A significant inverse relationship was also found between plasma n-3 polyunsaturated fatty acids and Lp(a) levels (rϭϪ0.24, Pϭ0.01). The results of this study are consistent with the concept that a diet rich in n-3 polyunsaturated fatty acids, and not genetic differences, is responsible for the lower plasma levels of Lp(a) in the fish-eating Bantus and strongly suggest that a sustained fish-based diet is able to lower plasma levels of Lp(a).