Dietary interventions with fish oil have been found to protect against the development of high-fat diet-induced insulin resistance and to decrease the expression of tumor necrosis factor (TNF)-alpha. However, the effect of fish oil administration on preexisting insulin resistance is subject to debate. In the present study, we examined the mechanism by which fish oil affects preexisting insulin resistance. High fat diet-induced insulin-resistant ApoE*3-Leiden transgenic mice were treated for 10 wk as follows: 1) high fat diet (control group), 2) high fat diet with 3 g/100 g fish oil and 3) high fat diet but food intake restricted to 75% of the ad libitum food intake. We measured plasma glucose, insulin, free fatty acids (FFA) and triglyceride (TG) levels throughout the study. After the 10-wk dietary intervention period we performed hyperinsulinemic euglycemic analyses and measured insulin sensitivity and FFA turnover. Furthermore, we then determined the VLDL-TG production rate and TNF-alpha protein expression in white adipose tissue (WAT). Compared with control mice, the insulin sensitivity of mice treated with fish oil was not affected, whereas it was improved (P < 0.05) for energy-restricted mice. FFA turnover was unaffected in both fish oil-treated and energy-restricted mice. Compared with controls, hepatic VLDL-TG production was lower (P < 0.05) with fish oil feeding but greater with energy restriction (P < 0.05). Interestingly, the level of TNF-alpha protein in WAT was lower (P < 0.05) in both groups. We conclude that partial replacement of saturated fat by fish oil does not improve preexisting high fat diet-induced insulin resistance, although it lowers TNF-alpha protein levels in WAT.