Effects of physiological increments of epinephrine (Epi) and cortisol (F) on glucose metabolism were assessed in dogs just before and during an intravenous glucose tolerance test performed in the last 3 h of an acute (short F + Epi, 4 h F and 3.5 h Epi) or prolonged (long F + Epi, 75 h F and Epi) infusion period. Comparison of the F + Epi effects with those of F and Epi alone enabled us to describe interactions between these hormones. The increase in plasma glucose after long F + Epi [from control (saline, Sal) of 5.2 +/- 0.1 to 5.8 +/- 0.1 mmol/l; n = 8; P less than 0.01] was not greater than the sum of the glucose increments after long F and long Epi individually. Long Epi and long F both reduced glucose tolerance (KGlc) significantly, but the decline during long F + Epi (from Sal 3.6 +/- 0.7 to 2.9 +/- 0.5%/min; P greater than 0.1) was less than during either individual infusion. Minimal model analysis showed that F attenuated the inhibitory effects of long Epi on glucose-mediated glucose disposal (SGlc), so that it was not reduced from 3.8 +/- 0.8 (Sal) during long F + Epi compared with the fall to 1.3 +/- 0.7 x 10(-2) min-1 (n = 6; P less than 0.05) during long Epi alone. F had the dominant influence on insulin sensitivity (SI) during infusion of F + Epi. The reduction of SI from 8.4 +/- 1.1 (Sal) to 6.6 +/- 1.2 (short F + Epi) and 5.1 +/- 1.1 x 10(-4) min-1 per mU/l (long F + Epi; P less than 0.05) paralleled that seen with F alone but contrasted with the acute reduction of SI during short Epi (4.8 +/- 1.5; P less than 0.02 vs. Sal) and its restoration to control values of 9.0 +/- 2.1 x 10(-4) min-1 per mU/l during long Epi. We conclude that Epi and F have distinct but nonadditive effects on determinants of glucose tolerance.