AICA riboside (0.1 to 1.0mM) caused a concentration-related increase of insulin output caused by D-glucose (5.6 to 20.0mM) in either rat isolated pancreatic islets or perfused pancreases. In the latter model, the rate of insulin release was further enhanced upon removal of AICA riboside from the perfusate. No insulinotropic action of AICA riboside was observed in the absence of D-glucose or at a low concentration (2.8mM) of the hexose. Preincubation of isolated islets for 30 min in the presence of AICA riboside (0.5 to 1.0mM) also enhanced insulin release recorded over 60 min incubation, in the absence of AICA riboside, but presence of either D-glucose (8.3mM), 2-ketoisocaproate (10.0mM), or the association of D-glucose (5.6 mM) and 2-ketoisocaproate (5.0 mM). The preincubation of the islets with AICA riboside failed, however, to augment the later secretory response to the association of L-leucine (10 mM) and either L-glutamine (10 mM) or L-asparagine (10 mM). In perfused pancreases exposed to 6 mM D-glucose, the presence of L-asparagine (10 mM) did not augment the insulinotropic action of AICA riboside. It is concluded that AICA riboside displays positive insulinotropic potential. However, the determinants of such an insulinotropic action remain to be elucidated.