The effect of fish consumption on plasma lipoprotein subfraction concentrations was studied in 22 men and women (age > 40 y). Subjects were provided an average American diet (AAD, 35% of energy as fat, 14% as saturated fat, and 35 mg cholesterol/MJ) for 6 wk before being assigned to a National Cholesterol Education Program (NCEP) Step 2 high-fish diet (n = 11, 26% of energy as fat, 4.5% as saturated fat, and 15 mg cholesterol/MJ) or a NCEP Step 2 low-fish diet (n = 11, 26% of energy as fat, 4.0% as saturated fat, and 11 mg cholesterol/MJ) for 24 wk. All food and drink were provided to study participants. Consumption of the high-fish NCEP Step 2 diet was associated with a significant reduction in medium and small VLDL, compared with the AAD diet, whereas the low-fish diet did not affect VLDL subfractions. Both diets significantly reduced LDL cholesterol concentrations, without modifying LDL subfractions. Both diets also lowered HDL cholesterol concentrations. However, the high-fish diet significantly lowered only the HDL fraction containing both apolipoprotein (apo) AI and AII (LpAI:AII) and did not change HDL subfractions assessed by NMR, whereas the low-fish diet significantly lowered the HDL fraction containing only apo AI (LpAI) and the large NMR HDL fractions, resulting in a significant reduction in HDL particle size. Neither diet affected VLDL and LDL particle size. Our data indicate that within the context of a diet restricted in fat and cholesterol, a higher fish content favorably affects VLDL and HDL subspecies.