Arterial concentrations and substrate exchange across the leg and splanchnic vascular beds were determined for glucose, lactate, pyruvate, glycerol, individual acidic and neutral amino acids, and free fatty acids (FFA) in six subjects at rest and during 4 h of exercise at approximately 30% of maximal oxygen uptake. FFA turnover and regional exchange were evaluated using (14)C-labeled oleic acid. The arterial glucose concentration was constant for the first 40 min of exercise, but fell progressively thereafter to levels 30% below basal. The arterial insulin level decreased continuously, while the arterial glucagon concentration had risen fivefold after 4 h of exercise. Uptake of glucose and FFA by the legs was markedly augmented during exercise, the increase in FFA uptake being a consequence of augmented arterial levels rather than increased fractional extraction. As exercise was continued beyond 40 min, the relative contribution of FFA to total oxygen metabolism rose progressively to 62%. In contrast, the contribution from glucose fell from 40% to 30% between 90 and 240 min. Leg output of alanine increased as exercise progressed. Splanchnic glucose production, which rose 100% above basal levels and remained so throughout exercise, exceeded glucose uptake by the legs for the first 40 min but thereafter failed to keep pace with peripheral glucose utilization. Total estimated splanchnic glucose output was 75 g in 4 h, sufficient to deplete approximately 75% of liver glycogen stores. Splanchnic uptake of gluconeogenic precursors (lactate, pyruvate, glycerol, alanine) had increased 2- to 10-fold after 4 h of exercise, and was sufficient to account for 45% of glucose release at 4 h as compared to 20-25% at rest and at 40 min of exercise. In the case of alanine and lactate, the increase in precursor uptake was a consequence of a rise in splanchnic fractional extraction. It is concluded that during prolonged exercise at a low work intensity (a) blood glucose levels fall because hepatic glucose output fails to keep up with augmented glucose utilization by the exercising legs; (b) a large portion of hepatic glycogen stores is mobilized and an increasing fraction of the splanchnic glucose output is derived from gluconeogenesis; (c) blood-borne substrates in the form of glucose and FFA account for a major part of leg muscle metabolism, the relative contribution from FFA increasing progressively; and (d) augmented secretion of glucagon may play an important role in the metabolic adaptation to prolonged exercise by its stimulatory influence on hepatic glycogenolysis and gluconeogenesis.