The pathogenesis of glucose intolerance in uremia was examined with the glucose clamp technique. Hyperglycemic clamp (n = 8): The plasma glucose concentration is acutely raised and maintained at 125 mg/dl above basal levels. Under these steady state conditions the glucose infusion rate, M, equals the amount of glucose metabolized: Predialysis M averaged 4.23 +/- 0.36 mg/kg/min and increased to 7.71 +/- 0.43 postdialysis (p less than 0.001). The plasma insulin response predialysis was 90 +/- 20 microU/ml and decreased to 80 +/- 23 microU/ml following dialysis. Consequently the M/l ratio, a measure of tissue sensitivity to insulin, increased by 80% +/- 25% (p less than 0.001) but still remained less than controls (p less than 0.01). Euglycemic insulin clamp (n = 10): The plasma insulin concentration is acutely raised by 100 microU/ml and the plasma glucose concentration is held constant at the basal level. Predialysis both M (3.37 +/- 0.36 mg/kg/min) and M/l (3.56 +/- 0.33 mg/kg/min per microU/ml X 100) were significantly less than controls (p less than 0.01). Postdialysis both M and M/l increased significantly (p less than 0.01) to a mean that was not significantly different from controls. Basal hepatic glucose production (n = 6), 2.15 +/- 0.09 mg/kg/min, was similar to controls and fell (87% +/- 4%) normally during the insulin clamp. In five uremic subjects in wom insulin binding to monocytes was measured, there was no correlation with tissue sensitivity to insulin (M/l). Significant abnormalities in both growth hormone and glucagon physiology were present in uremic individuals, but no correlation with either the presence or degree of glucose intolerance was demonstrable. In conclusion, glucose intolerance is universally present in uremic subjects and results primarily from peripheral tissue insensitivity to insulin. Insulin secretion is usually enhanced in an attempt to compensate for this insulin resistance but in occasional subjects uremia also inhibits beta cell sensitivity to glucose. Hepatic glucose production is unaffected by uremia. The lack of correlation between insulin binding and tissue sensitivity to insulin suggests that the cellular mechanism accounting for the insulin resistance is probably the result of a defect in intracellular metabolism or in the glucose transport system.