Isolated perifused rat islets were stimulated with glucose, exogenous insulin, or carbachol. C-peptide and, where possible, insulin secretory rates were measured. Glucose (8 -10 mM) induced dose-dependent and kinetically similar patterns of C-peptide and insulin secretion. The addition of 100 nM bovine insulin had no effect on C-peptide release in response to 8 -10 mM glucose stimulation. The addition of 100 nM bovine insulin or 500 nM human insulin together with 3 mM glucose had no stimulatory effect on C-peptide secretion rates from perifused rat islets. Stimulation with carbachol plus 7 mM glucose enhanced both C-peptide and insulin secretion, and the further addition of 100 nM bovine insulin had no inhibitory effect on C-peptide secretory rates under this condition. Perifusion studies using pharmacologic inhibitors (genistein and wortmannin) of the kinases thought to be involved in insulin signaling potentiated 10 mM glucose-induced secretion. The results support the following conclusions. 1) C-peptide release rates accurately reflect insulin secretion rates from collagenase-isolated, perifused rat islets. 2) Exogenously added bovine insulin exerts no inhibitory effect on release to several agonists including glucose. 3) In the presence of 3 mM glucose, exogenously added bovine or human insulin do not stimulate endogenous insulin secretion.Insulin secretion from the pancreatic -cell is tightly regulated by stimulatory signals generated during the intracellular metabolism of glucose and by neurohumoral agonists operative at the cell membrane (1-5). Most recently an additional layer of complexity has been added by reports suggesting that insulin exerts an autocrine stimulatory effect on insulin secretion from the -cell (6, 7). This concept was based primarily on amperometric measurements using -cells preincubated in 5-hydroxytryptamine (5HT).1 Because 5HT exposure exerts inhibitory effects on insulin release (8 -10), the precise physiologic significance of findings made with 5HT-preloaded -cells is unclear. Moreover, previous studies exploring the potential role of insulin on stimulated secretion showed no effect (11) or supported the concept that insulin exerts a negative, not positive, feedback on its own secretion (12-15).There are at least three major issues that must be addressed in attempting to establish the impact of exogenously added insulin on endogenous insulin secretory rates. The first is technical; it is difficult to accurately measure endogenous insulin release rates in the presence of exogenous insulin. The second relates to concentration of insulin necessary to establish an effect of exogenously added hormone on the -cell. Considering that the -cell continually releases insulin into a small volume of interstitial fluid, the level of insulin bathing the -cell may be quite high. For example, calculations based on islet cell volume, the insulin diffusion constant and insulin secretory rates, suggest that these levels may exceed 100 nM during glucose stimulation (16). Even at rest, levels o...