We and others have shown that insulin acutely suppresses glucose production in fasting nondiabetic humans and dogs, by both a direct hepatic effect and an indirect (extrahepatic) effect, and in diabetic dogs by an indirect effect alone. In type 2 diabetes, there is resistance to insulin's ability to suppress hepatic glucose production, but it has not previously been determined whether the resistance is primarily at the level of the hepatocyte or the peripheral tissues. To determine whether the diabetic state reduces the direct effect of insulin in humans, we studied nine patients with untreated type 2 diabetes who underwent three studies each, 4-6 weeks apart. 1) Portal study (POR): intravenous tolbutamide was infused for 3 h with calculation of pancreatic insulin secretion from peripheral plasma C-peptide. 2) Peripheral study (PER): equidose insulin was infused by peripheral vein. 3) Half-dose peripheral insulin study (1/2 PER): matched peripheral insulin levels with study 1. In all studies, glucose was clamped at euglycemia, glucose turnover was measured with the constant specific activity method, and 3-[3H]glucose was purified by high-performance liquid chromatography. Peripheral insulin was lower in POR versus PER but slightly higher in POR versus 1/2 PER, although most of the difference could be accounted for by higher proinsulin levels in POR (stimulated by tolbutamide). Calculated portal insulin was approximately 1.3-fold higher in POR versus PER and approximately 2.2-fold higher in POR versus 1/2 PER. In the final 30 min of the clamp, glucose production reached a lower steady-state level in PER than in POR (4.0 +/- 0.4 vs. 5.3 +/- 0.5 pmol(-1) x kg(-1) x min(-1), P < 0.05), despite the higher hepatic insulin level in POR. In contrast with our studies in nondiabetic individuals, glucose production was not more suppressed at steady state in POR versus 1/2 PER (5.3 +/- 0.4 micromol x kg(-1) x min(-1)), despite much higher hepatic insulin levels in POR. In conclusion, this is the first study in patients with type 2 diabetes to characterize insulin resistance to the acute direct suppressive effect of insulin on hepatic glucose production.