Tolbutamide (370 microM), gliclazide (62 microM) and glibenclamide (1 microM) failed to enhance insulin release evoked by 2-ketoisocaproate (10 mM) in rat pancreatic islets. Gliclazide also little affected insulin release evoked by 2-ketoisocaproate, whereas the hypoglycemic sulfonylurea stimulated insulin release from islets incubated in the absence of exogenous nutrient or presence of either L-glutamine, D-glucose, D-mannose, D-glyceraldehyde, L-leucine or the combination of D-glucose and pyruvate. In the presence of 2-ketoisocaproate, a modest secretory response to gliclazide was observed when the concentration of the 2-keto acid was decreased to 5 mM, or in perifused islets in which case gliclazide caused a transient increase in both 45Ca outflow and insulin output from prelabelled islets exposed to 10 mM 2-ketoisocaproate. Gliclazide and other hypoglycemic sulfonylureas failed to affect the oxidation of 2-[U-14 c]-ketoisocaproate and the latter 2-keto acid failed to affect the ionophoretic action of gliclazide in an artificial membrane model. Gliclazide increased 45Ca net uptake by islets exposed to 2-ketoisocaproate, but this effect of the sulfonylurea was much less marked than that seen in the presence of D-glucose used at a concentration of equal insulinotropic efficiency. These findings indicate that 2-ketoisocaproate impairs the cationic and secretory responses of islets to hypoglycemic sulfonylureas. It is proposed that such an impairment is compatible with the view that a remodelling of ionic fluxes in the islet cells represents a primary event in the process of sulfonylurea-stimulated insulin release.