Eight male non-insulin-dependent diabetic patients participated in a double-blind randomized cross-over study (2 weeks for each period) evaluating the effects of 10 g/day fish oil dietary supplementation on glucose and lipid metabolism. Fasting serum triglyceride concentrations were decreased by fish oil because of a reduction in VLDL (1.4 +/- 0.2 vs. 1.9 +/- 0.2 mmol/l, P less than 0.025). LDL cholesterol concentration was instead increased (3.4 +/- 0.3 vs. 2.8 +/- 0.3 mmol/l, P less than 0.025) and net changes in VLDL triglyceride and in LDL cholesterol were inversely correlated (r = -0.86, P less than 0.01). Plasma free fatty acids concentrations and turnover rate [( 3H]palmitate method) were similar after fish oil and placebo. Fish oil supplement did not induce significant changes in fasting blood glucose (8.1 +/- 1.1 vs. 8.5 +/- 1.2 mmol/l) and average daily blood glucose (BG) (9.4 +/- 3.2 vs. 9.3 +/- 3.5 mmol/l). Glucose stimulated plasma insulin response during a hyperglycemic clamp was not significantly influenced by fish oil both in the early phase and during steady state. Insulin sensitivity (M/I index) was also unchanged. In conclusion, this study shows that a dietary supplement of fish oil decreases plasma triglyceride levels in non-insulin-dependent diabetic patients, an increased conversion rate of VLDL to LDL playing a role in this change. With this dosage of fish oil no relevant variations in glycemic control, insulin secretion and insulin sensitivity occurred.