Pancreas transplantation is a controversial form of therapy for type 1 diabetes. A major obstacle to acceptance of this procedure for many physicians is the lack of demonstrable long-term success. We performed these studies to assess the hypothesis that successful pancreas transplantation is efficacious in normalizing endogenous insulin secretion and glycemia in the long term (1-2 decades). Sixteen patients with a history of diabetic complications who had undergone a transplant 10-18 years earlier involving either a whole or a segment of pancreas were recruited for measurements of fasting plasma glucose, HbA(1c), intravenous glucose tolerance, and insulin secretory reserve. All patients were taking immunosuppressive drugs, but none was using insulin or other hypoglycemic agents. All recipients had normal levels of fasting blood glucose, intravenous glucose tolerance, and HbA(1c), and 15 of 16 stated that their quality of life had improved after transplantation. They had intact acute insulin responses to intravenous pulses of glucose and to arginine and insulin secretory reserve. Glucose potentiation of arginine-induced insulin secretion, the measure of insulin secretory reserve, correlated significantly (r = 0.095, P < 0.001) with the acute insulin response to intravenous glucose, rendering the latter a much simpler and valid measure of functional β-cell mass. We conclude that successful pancreas transplants are efficacious for periods as long as 1-2 decades in returning euglycemia to type 1 diabetic patients by restoring endogenous insulin secretion and insulin secretory reserve. Thus, concern about long-term deterioration, as distinct from rejection, should not be a major obstacle when deciding whether to recommend pancreas transplantation.