A brief period of starvation (2-3 d) depletes the hepatic glycogen stores but results in only a limited reduction of the muscle glycogen depots. In this situation insulin resistance contributes to the glucose intolerance, but it is not known which tissue or tissues are responsible for the decreased insulin sensitivity. The present study was therefore undertaken to examine the influence of a 60-h fast on insulin sensitivity in splanchnic and peripheral tissues in normal humans. Euglycemic (95 mg/dl) 1-mU insulin and hyperglycemic (215-225 mg/dl) glucose clamp studies were conducted for 2 h in overnight (12 h) and prolonged (60 h) fasted nonobese subjects. Splanchnic exchange of glucose and gluconeogenic precursors was measured using the hepatic vein catheter technique.During the euglycemic clamp, insulin infusion resulted in similar steady state insulin levels in 60-h and 12-h fasted subjects (73±7 vs. 74±5 ttU/ml). Total glucose disposal was reduced by 45% after 60 h of fasting (4.0±03 vs. 7.6±1.1 mg/kg per min, P < 0.05) and the splanchnic glucose balance reverted from a net release in the basal state (12 h fast, -1.7±0.2, and 60-h fast, -0.9±0.1 mg/kg per min, P < 0.01) to a net uptake during the clamps that was similar after 60 h and 12 h of fasting (0.6±0.1 vs. 0.6±0.2 mg/kg per min).During the hyperglycemic clamp, insulin levels rose rapidly in all subjects. In the 12-h fasted group this rise was followed by a further gradual one, reaching significantly higher values than in 60-h fasted subjects during the second hour (67±15 vs. 25±2 AU/ml, P < 0.05). Total glucose disposal was lower, though not significantly so, after the 60-h fast (2.6±0.4 vs. 5.4±13 mg/kg per min, 0.05 < P < 0.10), and as with the euglycemic clamp, the splanchnic glucose balance was altered from a basal net release to a net uptake during the clamp (1.3±0.2 vs. 1.1±0.2 mg/kg per min).After an overnight fast, splanchnic lactate uptake fell and the arterial lactate concentration rose in response to both hyperglycemia and hyperinsulinemia, whereas these variables were unchanged in the 60-h fasted subjects during both types of clamp studies. We conclude that peripheral tissues are predominantly responsible for the insulin resistance seen in starvation. In contrast, net splanchnic glucose uptake in response to hyperinsulinemia and/or hyperglycemia remains unchanged during starvation, despite the depletion of the hepatic glycogen stores.