Serum lipoprotein[a] (Lp[a]) is a strong risk factor for coronary heart disease. We therefore examined the effect of dietary fatty acid composition on serum Lp[a] levels in three strictly controlled experiments with healthy normocholesterolemic men and women. In Expt. I, 58 subjects consumed a control diet high in saturated fatty acids for 17 days. For the next 36 days, 6.5% of total energy intake from saturated fatty acids was replaced by monounsaturates plus polyunsaturates (monounsaturated fatty acid diet; n = 29) or by polyunsaturates alone (polyunsaturated fatty acid diet; n = 29). Both diets caused a slight, nonsignificant, increase in median Lp[a] levels, with no difference between diets. In Expt. II, 10% of energy from the cholesterol-raising saturated fatty acids (lauric, myristic, and palmitic acid) was replaced by oleic acid or by trans-monounsaturated fatty acids. Each of the 59 participants received each diet for 3 weeks in random order. The median level of Lp[a] was 26 mg/l on the saturated fatty acid diet; it increased to 32 mg/l (P less than 0.020) on the oleic acid diet and to 45 mg/l (P less than 0.001) on the trans-fatty acid diet. The difference in Lp[a] between the trans-fatty acid and the oleic acid diets was also highly significant (P less than 0.001). Expt. III involved 56 subjects; all received 8% of energy from stearic acid, from linoleic acid, or from trans-monounsaturates, for 3 weeks each. All other nutrients were equal.(