Recent studies indicate that marine (omega-3) fatty acids decrease hypertriglyceridemia but worsen hyperglycemia in diabetes. We studied dose-response relationships between omega-3 intake and indices of carbohydrate and lipid metabolism in 21 hypertriglyceridemic patients with (n = 6) or without (n = 15) diabetes, and 6 normal volunteers. All subjects consumed isocaloric diets of 65% carbohydrate, 20% fat, and 15% protein. The basal diet contained 15% of total calories as vegetable oil (omega-6), and the test diets included 15%, 7.5%, or 3.75% calories as fish oil (MaxEPA). After three months of the basal diet, patients were randomized to receive two 3-month omega-3 diets in the following sequences: 15%/7.5%, 7.5%/15%, 7.5%/3.75%, or 3.75%/7.5%. Both 15% and 7.5% diets, regardless of sequence, significantly decreased serum triglycerides but increased low-density lipoprotein (LDL)-cholesterol levels as much as 98% and LDL/high-density lipoprotein (HDL)-cholesterol ratio as much as 1.6-fold. Daily insulin requirements of three diabetic patients increased progressively while they received an omega-3-enriched diet for up to 2 years. In healthy controls, favourable changes induced by an omega-3 fatty acid diet in serum lipids and lipoproteins were associated with a tendency toward an inhibition of C-peptide secretion following a meal challenge. We conclude that substitution of commercially available omega-3 for omega-6 fatty acids improves hypertriglyceridemia but may worsen other lipoproteins indices and may increase insulin requirements in diabetic hypertriglyceridemic subjects.