Hepatic glucose production and peripheral glucose utilisation were measured in vivo with the euglycaemic-hyper-insulinaemic clamp technique in rats rendered severely diabetic with streptozotocin (45 mg/kg) and in control rats. The rats were studied in the post-absorptive state while anaesthetised. The basal glucose production and glucose utilisation were significantly higher (p less than 0.001) in diabetic rats 9 days after streptozotocin administration. During the clamp studies, suppression of glucose production by the liver induced by submaximal or maximal insulin levels was significantly less (p less than 0.01 and p less than 0.001 respectively) effective in diabetic rats as compared to control rats. Glucose utilisation was significantly lower following both submaximal (p less than 0.01) or maximal (p less than 0.001) hyperinsulinaemia as compared to control rats. Oral administration of vanadate (0.2 mg/ml in drinking water) for a 20-day period in diabetic rats lowered their plasma glucose levels to normal near values within 4 days, normalised plasma insulin levels, and increased pancreatic insulin stores. The rate of glucose disappearance (K value) and in vivo glucose-induced insulin secretion as estimated during an i.v. glucose tolerance test were not significantly improved. In control rats, vanadate treatment did not significantly affect any of the above parameters. In vanadate treated diabetic rats, basal glucose production was normalised. Following submaximal or maximal hyperinsulinaemia, glucose production was suppressed normally. Basal glucose utilisation was restored and returned to normal values during submaximal hyperinsulinaemia. However, during maximal hyperinsulinaemia, glucose utilisation still remained significantly lower (p less than 0.05) as compared to vanadate-treated control rats. (ABSTRACT TRUNCATED AT 250 WORDS)