It is not yet clear whether a low-fat or a high-fat diet is more suitable for the treatment of hypertriglyceridemia. Therefore, we conducted a dietary study with nonobese hypertriglyceridemic men using a randomized crossover design. After a 2-wk acclimation period, the subjects were randomly assigned to 2 groups. One group consumed a low-fat (29% of energy) diet for 3 wk, followed first by a 2-wk washout period, then a 3-wk high-fat (40% of energy) diet period. The second group consumed the low- and high-fat diets in reverse order. Both diets were isocaloric, rich in monounsaturated fatty acids (MUFAs) and long-chain (n-3)-PUFAs, and contained no alcohol. Absolute amounts of long-chain (n-3)-PUFAs and fiber were similar, as were the fatty acid and carbohydrate compositions. Serum triglyceride concentrations in fasting subjects decreased during the high-fat diet period by 34% (95% CI = -19 to -46%, P = 0.001), and during the low-fat diet period by 31% (95% CI = -0.5 to -51.8%, P = 0.048). Triglyceride concentrations did not differ between the low- and high-fat diet periods. However, the high-fat diet lowered triglyceride concentrations more effectively in all subjects with a baseline triglyceride concentration < 4.5 mmol/L, whereas the low-fat diet lowered triglyceride levels more effectively in half of the subjects with a baseline value > 4.5 mmol/L. Based on these findings, we recommend a high-fat triglyceride-lowering diet for patients with only slightly elevated serum triglyceride concentrations (<4.5 mmol/L). However, a lower-fat diet is more suitable for some subjects with more distinctly elevated triglyceride concentrations.