Summary. Glucose-induced insulin secretion is lost in the face of chronic hyperglycaemia. The same defect is present when normal rats are made hyperglycaemic by 48-h glucose infusions. Insulin secretory responses were mapped out during the post-infusion period in order to determine how long it takes for normal Beta-cell function to recover, and to identify factors which influence the rate of recovery. Male Sprague Dawley rats weighing 200-250 g were infused with 50% glucose or 77 mmol/1 NaC1 for 48 h. The glucose-infused rats were mildly hypoglycaemic for 14 h after the infusion ceased. Glucose-induced insulin secretion, absent at the end of the glucose infusion, was normal 6 h post-infusion. Such rapid recovery was not because of the short duration of hyperglycaemia; mild hypoglycaemia from a 5-h insulin infusion in 90% pancreatectomized rats resulted in a four-fold rise in glucose-induced insulin secretion. Under in vitro conditions, extreme glucose deprivation caused by perfusing the pancreas of glucose-infused rats with buffer devoid of glucose restored glucose-induced insulin secretion in just 37 min. Therefore, the suppression of glucose-induced insulin release by chronic hyperglycaemia is a dynamic situation that requires ongoing hyperglycaemia to prevent the reappearance of glucose responsiveness. This study shows recovery of glucose-induced insulin secretion after just 6 h of mild hypoglycaemia in vivo and even faster recovery with more severe glucose deprivation in vitro. Our results suggest that there is an inverse relationship between the rate of return of Betacell glucose responsiveness and the ambient glucose concentration.Key words: Animal models, Type 2 (non-insulin-dependent) diabetes mellitus, insulin secretion, insulin treatment, hypoglycaemia, islets of Langerhans.Normally, the plasma glucose concentration exerts the dominant influence over insulin secretion. Type 2 (noninsulin-dependent) diabetes mellitus is characterized by Beta cells which no longer respond to a changing glucose level [1][2][3][4]. We have proposed that glucose-induced insulin secretion is lost as a direct result of Beta cells being exposed to a chronically elevated glucose concentration [4,5]. Support for this idea has come from animal studies in which normal rats made hyperglycaemic through a variety of mechanisms develop a similar defect [6.9], and also that normalizing the plasma glucose concentration in these rats with phloridzin restores glucose-induced insulin secretion [10,11].Little is known about reversal of secretory defects in non-insulin-dependent diabetes mellitus except that restoration of euglycaemia often results in the reappearance of glucose-induced insulin release [4, 12-14]. The recovery time has not been mapped out, nor has much been learned about what conditions speed up or retard recovery. Animal models should theoretically provide a way to investigate this question, but reversal studies in hyperglycaemic rat models have provided conflicting results. Using in vitro techniques, glucose-induced insuli...