Endogenous glucose production has been shown to increase during administration of glucagon + fructose, but not during administration of fructose alone. To determine the mechanisms by which glucagon exerts this action, endogenous glucose production (EGP) and gluconeogenesis from fructose (GNF) were measured in eight healthy subjects infused 1) with graded doses of glucagon (2 and 4 ng.kg-1.min-1 for 3 h each) during constant infusion of 13C-fructose (3 mg.kg-1.min-1), and 2) with graded doses of 13C-fructose (3 and 6 mg.kg-1.min-1) during constant glucagon infusion (2 ng.kg-1.min-1). GNF was estimated from 13C-glucose synthesis. In both protocols, infusion of 3 mg.kg-1.min-1 fructose + 2 ng.kg-1.min-1 glucagon increased EGP by 5-8% (P < 0.05), while GNF represented 43-49% of EGP. Thereafter, increasing the glucagon infusion rate further increased EGP to 118 +/- 3% of basal values (P < 0.01) without altering the proportion due to GNF. In contrast, increasing the fructose infusion rate at constant glucagonemia increased EGP similarly (by 19 +/- 4%, P < 0.05) but enhanced the contribution of GNF to 76 +/- 2% (P < 0.001). Graded infusion of glucagon or fructose alone failed to stimulate EGP. The present findings indicate that hyperglucagonemia stimulates endogenous glucose production during fructose infusion. This effect is not secondary to a stimulation of gluconeogenesis, but to a channelling of glucose-6-phosphate towards systemic release.