This study was undertaken to assess the mechanism by which prostaglandins of the E series inhibit glucose-induced insulin secretion in man. Acute insulin response (mean change 3\p=n-\10 min) to iv glucose (0.33 g/kg) was decreased by 40% during the infusion of prostaglandin E2 (10 \g=m\g/min) and glucose disappearance rates were reduced (P <0.05). Insulin response to arginine (5 g iv) and tolbutamide (1 g iv) were not affected by the same rate of prostaglandin E2 infusion. The inhibitory effect of prostaglandin E2 on glucoseinduced insulin secretion was prevented by theophylline (100 mg as a loading dose followed by a 5 mg/min infusion), a drug that increases the intracellular cAMP concentrations by inhibiting phosphodiesterase activity. Our data suggest the involvement of the adenylate cyclase system in the inhibitory action of prostaglandin E2 on glucose-induced insulin secretion in man.The initial confusion about the effects of prostaglandins (PG) on insulin secretion, arising for the main part from the conflicting results of the in vitro studies performed in different animal spe¬ cies (Robertson 1979), was followed by an unceas¬ ing increase in experimental in vivo data indicat¬ ing that PG of the E series (PGE) inhibit the insulin response to glucose in several species, including rats (D'Onofrio et al. 1977; Sacca et al. 1975), dogs (Robertson et al. 1974) and humans (Giugliano et al. 1978; Robertson & Chen 1977). The available data in humans indicate that both PGE and PGE2 can inhibit glucose-induced insu¬ lin secretion in a dose-dependent manner (Giug¬ liano & Torella 1978; Giugliano et al. 1978; Ro¬ bertson & Chen 1977) and that this effect is probably direct, because it is not reversed by alpha-adrenergic antagonists, which block the in¬ hibiting effect of catecholamines on insulin se¬ cretion (Giugliano et al. 1979). More recently, Giugliano et al. (1983) showed that endogenous, pancreatically produced PGE may play a role in the appearance of the typical biphasic pattern of insulin release following a square wave glucose stimulation in normal man. Despite this evidence, many important questions still remain and await answers. The more pressing of these questions concerns the mechanism of action of PGE in islet cells. Recent in vitro studies (Robertson et al. 1987) suggest that PGE may interfere with the adenylate cyclase in the pancreatic beta-cell. Ac¬ cordingly, the present study was undertaken to evaluate the effect of theophylline on PGE-mediated inhibition of insulin secretion in man, this drug inhibiting hydrolysis of cAMP by phospho¬ diesterase resulting in increased levels of intra¬ cellular cyclic nucleotides (Turtle et al. 1967).