Plasma glucose, C-peptide, and insulin responses to intravenous glucose (intravenous glucose tolerance test [IVGTT], 0.5 g/kg), glucagon (1 mg i.v.), and oral glucose (oral glucose tolerance test [OGTT], 1 g/kg) were assessed in six normal beagles before, during, and 1 and 4 mo after the administration of cyclosporin A (CsA) in doses previously shown to be required for uniform prevention of canine islet-allograft rejection (20 mg/kg; mean trough radioimmunoassay serum levels greater than or equal to 500 ng/ml). Insulin secretion in response to intravenous glucose and glucagon was significantly inhibited during the administration of CsA (areas under insulin-response curves, pmol.min-1.L-1; IVGTT, pre-CsA, 11,127 +/- 1285; during CsA, 5954 +/- 1147, P less than .05; glucagon tolerance test, pre-CsA, 18,617 +/- 2807; during CsA, 4401 +/- 486, P less than .05 vs. pretreatment levels). These secretory defects persisted 4 mo after CsA was discontinued (IVGTT, 4358 +/- 659; glucagon tolerance test, 10,567 +/- 2479, P less than .05). C-peptide responses paralleled these changes. Plasma glucose disposal in response to these secretagogues, however, returned to normal 1 mo after discontinuation of CsA. In contrast to the findings for IVGTT and glucagon, insulin-response curves to OGTT were not statistically different during CsA administration. We conclude that, although glucose disappearance rates are normal after discontinuation of the CsA administration, CsA causes irreversible impairment in islet secretory responses detectable with IVGTT and glucagon but not with OGTT. These results suggest that short-term CsA in doses required to prevent islet-allograft rejection in dogs can result in permanent loss of functionally competent beta-cells.