T he dose as well as the time kinetics of insulin and adenosine-3', 5' -monophosphate (cyclic AMP) responses to glucose were compared in pancreatic islets of fed and starved rats. There was a preferential impairment of the early phase of glucose-induced insulin release in perifused islets of rats starved for 16 and 48 h. Similarly, the accumulation of 3H cyclic AMP in islets prelabeled with 3H-2-adenine was less in islets of 48 h starved than fed rats, during the first 10-min of stimulation with 26.7 mM glucose in the presence of 0.1 mM of the phosphodiesterase inhibitor, 3-isobutyl-1-methylxanthine, whereas at 30 and 60 min 3H cyclic AMP responses to glucose were similar in fed and starved islets. Also, in 10-min incubations with glucose 3.3, 6.7, 10.0, 13.3, and 26.7 mM without and with 0.1 mM and 1.0 mM 3-isobutyl-1-methylxanthine, insulin release correlated strongly with the accumulation of 3H cyclic AMP in the islets of fed as well as starved rats. The thresholds for glucose-induced insulin and 3H cyclic AMP responses were higher and the maximal responses were lower in starved than fed islets. Preincubation of islets of 48-h starved rats with 16.7 mM glucose for 60 min corrected the impaired insulin and 3H cyclic AMP responses to glucose. Starvation-induced impairment of insulin secretory responses to glucose, and their restoration by preincubation with glucose in vitro, may represent acute regulatory effects of glucose on the adenylate cyclase-cyclic AMP system in the pancreatic beta cell.