Insulin inhibits its own release (autofeedback), and growth hormone (GH) inhibits the GH response to a variety of stimuli. The aim of this study was to evaluate whether glucagon (G) can modify pancreatic G (IRG) release in humans. Seven healthy men received intravenous (i.v.) arginine (30 g in 30 minutes) 240 minutes after the beginning of a 0.9% NaCI saline infusion and a 2.5-, 4.0-, and 8.0-ng/kg.min-1 porcine G infusion, with each infusion lasting 360 minutes. All G infusions yielded stable and dose-related plasma IRG levels, and the 4.0- and 8.0-ng/kg.min-1 G infusions decreased plasma free fatty acids (FFA) and blood glycerol and beta-OH-butyrate levels and elicited insulin (IRI) release, and the 8.0-ng/kg.min-1 G infusion elicited GH release and increased blood glucose (BG) levels; somatostatin (SRIF) levels were not affected by G infusions. At 240 minutes, plasma IRG levels were higher during G infusion than during saline infusion, whereas serum IRI and BG levels had returned to preinfusion levels. At this point, G infusions decreased the integrated (240 to 300 minutes) IRG, IRI, BG, and SRIF responses, but not the GH response to arginine. These data indicate that prolonged G infusions decrease the IRG response to arginine; in addition, G decreases plasma FFA levels, and higher G doses stimulate IRI release and exert a self-limited hyperglycemic effect. The fact that the IRI response to arginine was decreased by G could be due to a refractoriness of beta cells to subsequent stimuli; the decreased SRIF response to arginine is likely due to G itself or to a decrease of plasma FFA levels.