The analysis of mass isotopomers in blood glucose and lactate can be used to estimate gluconeogenesis (Gneo), glucose production (GP), and, by subtraction, nongluconeogenic glucose release by the liver. At 6 AM, 18 normal subjects received a 7-hour primed constant infusion of [U-13C6] glucose. After a 3-hour baseline period (12 hours of fasting), somatostatin, insulin, hydrocortisone, growth hormone (GH), and glucagon were infused for 4 hours. Glucagon was infused at a low-dose (n = 6) or high-dose (n = 6) concentration for 4 hours and was compared with fasting alone (n = 6). Low-dose glucagon infusion increased plasma glucagon (64 +/- 3 v 44 +/- 7 ng/L, low glucagon v baseline). GP increased above baseline (15.5 +/- 0.5 v 13.8 +/- 0.5 micromol/kg/min, P < .05), which was also greater than fasting alone (11 .5 +/- 0.6 micromol/kg/min, P < .05). The elevation in GP was due to a near doubling of nongluconeogenic glucose release compared with fasting alone (8.3 +/- 0.6 v 4.7 +/- 0.5 micromol/kg/min, P < .01). High-dose glucagon infusion (125 +/- 25 ng/L) increased GP above baseline (15.8 +/- 0.6 v 13.5 +/- 0.5 micromol/kg/min, P < .05), which was also greater than fasting alone (11.5 +/- 0.6 micromol/kg/min, P < .05). The increase in GP was due to an increase in Gneo (8.5 +/- 0.5 v 6.8 +/- 0.7 micromol/kg/min, P < .05) and nongluconeogenic glucose release (7.4 +/- 0.5 v 4.7 +/- 0.4 micromol/kg/min, P < .05) compared with fasting. Low-dose glucagon increases GP only by stimulation of nongluconeogenic glucose release. High-dose glucagon increases GP by an increase in both Gneo and nongluconeogenic glucose release.