The mechanisms of the hypocholesterolemic effect of polyunsaturated fats (PUSF) are not well understood. One possibility is that these fats uniquely reduce the cholesterol content of lipoproteins. The present study was carried out to determine specifically whether the ratio of cholesterol-to-protein (or apoB) in LDL (or other lipoproteins) is reduced by PUSF; also, lipoprotein composition was examined for other possible changes. Eight men and two women with different levels of plasma cholesterol were studied on the metabolic ward for 8 weeks. They were given a diet high in saturated fats (SF) for 4 weeks and another rich in PUSF for 4 weeks. On PUSF diets, mean plasma cholesterol decreased by 25% (SF = 296 +/- 27 (SEM) vs. PUSF = 223 +/- 21 mg/dl) as did total plasma apoB (155 +/- 8 vs. 116 +/- 13 mg/dl). LDL-Cholesterol decreased by 26%, and LDL-apoB fell by 29%. The mean ratio of cholesterol-to-apoB did not change significantly (SF = 1.52 +/- 0.04 vs. PUSF = 1.48 +/- 0.07). Likewise, HDL-cholesterol decreased by 15% (SF = 51 +/- 5 vs. PUSF = 43 +/- 4 mg/dl), and total plasma apoA-I was reduced by 19% (95 +/- 15 vs. 77 +/- 6 mg/dl); also, no change in the cholesterol-to-apoA-I in HDL was noted. Finally, there were no changes in cholesterol/apoB or triglyceride/apoB ratios in VLDL despite a 23% decrease in plasma triglycerides on PUSF. Thus, the hypocholesterolemic effect of PUSF was uniform for all lipoproteins and usually was accompanied by a corresponding decrease in concentrations of apoprotein constituents.