High levels of some but not all dietary fats lead to insulin resistance in rats. The aim of this study was to investigate the important determinants underlying this observation. Insulin action was assessed with the euglycemic clamp. Diets high in saturated, monounsaturated (omega-9), or polyunsaturated (omega-6) fatty acids led to severe insulin resistance; glucose infusion rates [GIR] to maintain euglycemia at approximately 1000 pM insulin were 6.2 +/- 0.9, 8.9 +/- 0.9, and 9.7 +/- 0.4 mg.kg-1. min-1, respectively, versus 16.1 +/- 1.0 mg.kg-1.min-1 in chow-fed controls. Substituting 11% of fatty acids in the polyunsaturated fat diet with long-chain omega-3 fatty acids from fish oils normalized insulin action (GIR 15.0 +/- 1.3 mg.kg-1.min-1). Similar replacement with short-chain omega-3 (alpha-linolenic acid, 18:3 omega 3) was ineffective in the polyunsaturated diet (GIR 9.9 +/- 0.5 mg.kg-1.min-1) but completely prevented the insulin resistance induced by a saturated-fat diet (GIR 16.0 +/- 1.5 mg.kg-1.min-1) and did so in both the liver and peripheral tissues. Insulin sensitivity in skeletal muscle was inversely correlated with mean muscle triglyceride accumulation (r = 0.95 and 0.86 for soleus and red quadriceps, respectively; both P less than 0.01). Furthermore, percentage of long-chain omega-3 fatty acid in phospholipid measured in red quadriceps correlated highly with insulin action in that muscle (r = 0.97). We conclude that 1) the particular fatty acids and the lipid environment in which they are presented in high-fat diets determine insulin sensitivity in rats; 2) impaired insulin action in skeletal muscle relates to triglyceride accumulation, suggesting intracellular glucose-fatty acid cycle involvement; and 3) long-chain omega-3 fatty acids in phospholipid of skeletal muscle may be important for efficient insulin action.