SummaryIn order to evaluate the role of portal insulin in the modulation of hepatic glucose production (HGP), measurements of plasma glucose and insulin concentrations and both HGP and peripheral glucose disappearance rates were made following an infusion of a dose of tolbutamide (0.74 mgm -2. min -1) in healthy volunteers that does not result in an increase in peripheral vein insulin concentrations or metabolic clearance rate of glucose. The results showed that the infusion of such a dose of tolbutamide was associated with a significant and rapid decline in both HGP (from 9.0 + 0.5 to 7.7 + 0.5~tmol.kg -1.min -1 or A= -13.8+4.5%; p< 0.001 compared to saline) and plasma glucose concentration (from 5.1+ 0.2 to 4.4+ 0.1mmol/l or A = -13.0 + 2.1%; p < 0.01 compared to saline). Since neither HGP nor fasting glucose fell when tolbutamide-stimulated insulin secretion was inhibited by the concurrent administration of somatostatin, it indicated that tolbutamide by itself, does not directly inhibit HGP. Finally, HGP fell by 26.3 + 6.0 % at 10 min after a dose of tolbutamide that elevated both peripheral and portal insulin concentrations, at a time when HGP had barely increased (A = + 6.9 + 5.3 %). The difference in the magnitude of the two responses was statistically significant (p < 0.03), providing further support for the view that insulin can directly inhibit HGP, independent of any change in flow of substrates from periphery to liver. [Diabetologia (1997[Diabetologia ( ) 40: 1300[Diabetologia ( -1306 Keywotds Liver, insulin, portal system, C-peptide, tolbutamide.Evidence has been published showing that hepatic glucose production (HGP) is decreased following the infusion of either insulin or sulfonylurea compounds [1][2][3][4], and this is generally assumed to be due to a direct effect of insulin on the liver. However, the results of two recent studies in dogs [5,6] of insulin, suggesting that the ability of insulin to inhibit HGP may be mediated through its peripheral action on substrate flow to the liver. Although we think it reasonable that insulin can inhibit HGP by decreasing substrate flow to the liver, the experiments to be presented were initiated to see whether insulin can also inhibit HGP directly. In order to accomplish this task non-invasively, we measured HGP, in normal subjects, in response to a continuous infusion of tolbutamide that would not significantly increase peripheral insulin concentrations and compared it to a saline) infusion. Under these conditions, HGP should have been similar in both studies if HGP was only modulated by the peripheral inhibition of substrate flow. Since HGP was significantly lower following tolbutamide, the data are consistent with the hypothesis that insulin can directly inhibit HGE