Twelve healthy men were fed in turn three diets for 6 wk each in a 3 × 3 randomized block design: a control diet (essentially fish free), a fish diet (200 g lean Australian fish flesh/d), and the same fish-based meal but supplemented with 5 g fish oil/d. Dietary eicosapentaenoic acid [EPA, 20: 5n-3 (ω-3)] was strongly associated with erythrocyte membrane EPA (r = 0.908 at 6 wk), strengthening its value as a measure of compliance in fish and fish-oil feeding trials. On the fish diet, subjects had increased incorporation of n-3 polyunsaturated fatty acids (PUFAs) at the expense of n-6 PLJFAs in their erythrocyte membranes. When the fish-based diet was supplemented with fish oil (5 g/d), there was a significant lowering of plasma triacylglycerol (-0.16 ± 0.24 mmol/L; x̄ ± SD). No change in plasma total cholesterol was detected although the fish+oil diet produced a reduction in very-low-density-lipoprotein cholesterol (-0.24 ± 0.26 mmol/L).