Glucose utilization was assessed before and after operation in five patients with insulin-secreting tumors using the euglycemic clamp. Two groups of age, sex, and weight-matched controls were studied under conditions of either acute (3 h, N = 7) or chronic (48-72 h, N = 6) hyperinsulinemia (46 +/- 4 microU/mL). The rate of glucose infusion (M = mg/kg/min) required to maintain steady-state euglycemia was taken as index of glucose utilization. M was higher both in postoperative patients and in chronic controls than in preoperative patients (2.1 +/- 0.1 and 1.8 +/- 0.2 vs. 1.0 +/- 0.2, p less than 0.01), yet never achieved levels seen in acute controls (3.3 +/- 0.3, p less than 0.01). Also reported is one subject who had a predominantly proinsulin-producing tumor with little insulin secretion, who also manifested a similar defect in glucose utilization. Excision of this tumor resulted in partial reversal of the peripheral defect in glucose metabolism. In conclusion, chronic endogenous hyperinsulinemia (and hyperproinsulinemia) in humans is associated with a defect in peripheral glucose utilization that is only partially resolved 2 years after tumor removal.