Arterial (A) and renal venous (RV) concentrations and net splanchnic exchange of glucose, fructose, lactate, pyruvate, glycerol, and alanine were studied in the basal state and during a 135-min intravenous infusion of fructose at 2 mmol/min in healthy subjects after a 60-h fast. After 45 min of the fructose infusion, somatostatin (9 microgram/min) was infused for 60 min to induce hypoglucagonemia. Fructose infusion resulted in a net uptake of this hexose by the kidney as well as the splanchnic bed. Estimated renal uptake of fructose could account for the disposal of 20% of the administered fructose load while splanchnic uptake accounted for 38%. The fructose infusion resulted in a rise in blood glucose of 0.9 mmol/L, a 35% increase in net glucose output from the splanchnic bed, and a consistent net output of glucose from the kidney (A-RV = -0.17 +/- 0.05 mmol/L as compared with 0 +/- 0.03 in the basal state, P less than 0.02). Net glucose release from the kidney could account for 55% of the net renal uptake of fructose. The fructose infusion also resulted in a marked change in renal lactate balance from a net uptake in the basal state (A - RV = 0.05 +/- 0.01 mmol/L) to a net output during fructose administration (A - RV = -0.10 +/- 0.04). Administration of somatostatin resulted in a fall in arterial glucagon levels and a 35% decrease in splanchnic glucose output but failed to alter the arterial-renal venous difference for glucose observed during the fructose infusion. We conclude that in 60-h fasted man: (a) intravenous infusion of fructose results in a net uptake of this hexose by the kidney as well as the liver, (b) this uptake is accompanied by stimulation of renal as well as hepatic glucose production and renal production of lactate, and (c) hypoglucagonemia inhibits splanchnic but not renal glucose output during fructose infusion. These data indicate that the kidney is an important site of fructose disposal and that glucose and lactate are end products of renal fructose metabolism.