It has long been assumed that replacement of insulin in insulin-deficient diabetic patients will normalise glucose utilization. In this study, glucose utilization was measured in nine long-standing, poorly controlled diabetic patients and five control subjects, matched for age (33 +/- 3 versus 33 +/- 2 years) and ponderal index (22.9 +/- 1.3 versus 21.7 +/- 1.0). Glucose uptake was measured during steady state insulinaemia in the diabetic patients and control subjects, at euglycaemia (5.5 +/- 0.5 versus 5.4 +/- 0.3 mmol/l, respectively) and moderate hyperglycaemia (11.8 +/- 0.9 versus 10.2 +/- 0.7 mmol/l, respectively). At euglycaemia with similar free insulin levels (50 +/- 19 versus 43 +/- 9 mU/l; p greater than 0.6), the diabetic patients utilized less glucose than the control subjects (27.8 +/- 4.2 versus 56.4 +/- 5.7 mumol.kg-1.min-1;.p less than 0.005). During hyperglycaemia, the diabetic patients utilized almost as much glucose as the control subjects did at euglycaemia (49.9 +/- 6.4 versus 56.4 +/- 5.7 mumol.kg-1.min-1, respectively). In the control subjects, a 1-mmol/l rise in glucose concentration (with insulin remaining constant) resulted in a 12.3 +/- 1.3 mumol.kg-1.min-1 rise in glucose utilization. In contrast, in the diabetic patients, a 1-mmol/l rise in blood glucose resulted in a rise in glucose utilization of only 3.8 +/- 0.8 mumol.kg-1.min-1 (p less than 0.001), in the presence of similar concentrations of plasma insulin. This defect of glucose utilization in Type1 diabetic patients could not be reversed by acutely raising insulin to 247 +/- 23 mU/l.(ABSTRACT TRUNCATED AT 250 WORDS)