The ability of hyperglycemia per se to suppress endogenous glucose production (GP) is blunted in type 2 diabetes. This could be due in part to decreased glucose-induced flux through glucokinase (GK). Because fructose activates hepatic GK, we examined whether catalytic amounts of fructose could restore inhibition of GP by hyperglycemia in humans with type 2 diabetes. Glucose fluxes ([3-(3)H]glucose) were measured during euglycemia (5 mmol/l) and after abrupt onset of hyperglycemia (10 mmol/l; variable dextrose infusion) under fixed hormonal conditions (somatostatin infusion for 6 h with basal insulin/glucagon/growth hormone replacement). A total of 10 subjects with moderately controlled type 2 diabetes and 7 age- and BMI-matched nondiabetic subjects were studied on up to three separate occasions under the following conditions: without fructose (F(-)) or with infusion of fructose at two dosages: 0.6 mg/kg center dot min (low F) and 1.8 mg/kg center dot min (high F). Although GP failed to decrease in response to hyperglycemia in type 2 diabetes, the coinfusion of both doses of fructose was associated with comparable decreases in GP in response to hyperglycemia (low F = -27%, high F = -33%; P < 0.01 vs. F(-) at both dosages), which approached the 44% decline in GP observed without fructose in the nondiabetic subjects. GP responses to hyperglycemia were not altered by the addition of fructose in the nondiabetic group (low F = -47%, high F = -42%; P > 0.05 vs. F(-)). Thus, the administration of small amounts of fructose to type 2 diabetic subjects partially corrected the regulation of GP by hyperglycemia per se, yet did not affect this regulation in the nondiabetic subjects. This suggests that the liver's inability to respond to hyperglycemia in type 2 diabetes, likely caused by impaired GK activity, contributes substantially to the increased GP in these individuals.