To determine the mechanism for cortisol enhancement of glucagon-stimulated overall hepatic glucose output (OHGO), we employed the glucose-insulin clamp technique with infusions of [6-3H]glucose and [U-14C]lactate and measured OHGO, glucose utilization, and the turnover and incorporation of lactate in plasma glucose in normal volunteers under four experimental conditions: 1) normoglucagonemia (approximately 150 pg/ml)- normocortisolemia (approximately 14 micrograms/dl); 2) isolated hyperglucagonemia (approximately 550 pg/ml); 3) isolated hypercortisolemia (approximately 32 micrograms/dl); and 4) combined hyperglucagonemia-hypercortisolemia. Isolated hyperglucagonemia caused initial increases in OHGO and lactate gluconeogenesis, which were maximal at 1 h (23.9 +/- 1 and 2.7 +/- 0.4 mumol.kg-1.min-1, respectively) but remained significantly above values in control experiments through 5 h (10.3 +/- 0.7 vs. 8.2 +/- 1.1, P less than 0.03; 2.2 +/- 0.4 vs. 1.2 +/- 0.3, mumol.kg-1.min-1, P less than 0.04, respectively). Hypercortisolemia has no effect on OHGO but increased lactate gluconeogenesis after 3 h. Superimposition of hypercortisolemia on hyperglucagonemia did not further increase OHGO (11.1 +/- 0.7 vs. 10.3 +/- 0.7 mumol.kg-1.min-1, P = NS) but augmented lactate gluconeogenesis additively (isolated hyperglucagonemia = 0.96, isolated hypercortisolemia = 0.98; combined = 2.02 mumol.kg-1.min-1). Neither glucagon nor cortisol affected lactate turnover or glucose utilization. We conclude that glucagon has a persistent effect on OHGO largely accounted for by increased gluconeogenesis. Cortisol augments glucagon-stimulated gluconeogenesis in an additive manner best explained by changes in gluconeogenic enzymes rather than in substrate availability. Finally, the fact that cortisol increased gluconeogenesis without affecting glucose utilization suggests that the liver is more sensitive to the diabetogenic effects of cortisol than are peripheral tissues.