Prolonged infusions of lipid and heparin that achieve high physiological free fatty acid (FFA) concentrations inhibit hepatic (and peripheral) insulin sensitivity in humans. These infusions are composed largely of polyunsaturated fatty acids (PUFA; linoleic and linolenic). It is not known whether fatty acid composition per se affects hepatic glucose metabolism in humans. To address this issue, we examined the impact of enteral infusions of either palm oil (48% palmitic, 35% oleic, and 8% linoleic acids) or safflower oil (6% palmitic, 12% oleic, 74% linoleic acids) in 14 obese nondiabetic subjects. ²H₂O was administered to determine the contribution of gluconeogenesis to endogenous glucose production (EGP), and a primed continuous infusion of [6,6-²H]glucose was administered to assess glucose appearance. As a result of the lipid infusions, plasma FFA concentrations increased significantly in both the palm oil (507.5 ± 47.4 to 939.3 ± 61.3 μmol/l, P < 0.01) and safflower oil (588.2.0 ± 43.0 to 857.8 ± 68.7 μmol/l, P < 0.01) groups after 4 h. EGP was similar at baseline (12.4 ± 1.8 vs. 11.2 ± 1.0 μmol·kg FFM⁻¹·min⁻¹). During a somatostatin-insulin clamp, the glucose infusion rate was significantly lower (AUC glucose infusion rate 195.8 ± 50.7 vs. 377.8 ± 38.0 μmol/kg FFM, P < 0.01), and rates of EGP were significantly higher (10.7 ± 1.4 vs. 6.5 ± 1.5 μmol·kg FFM⁻¹·min⁻¹, P < 0.01) after palm oil compared with safflower oil, respectively. Baseline rates of gluconeogenesis and glycogenolysis were also similar. However, after lipid infusion, rates of glycogenolysis were suppressed by safflower oil but not by palm oil. Thus these studies demonstrate, for the first time in humans, a differential effect of saturated fatty acids and PUFA on hepatic glucose metabolism.