Considerable evidence supports the hypothesis that the microangiopathic lesions of diabetes are secondary to disordered metabolism. This hypothesis underlies the rationale for whole pancreas and islet transplantation. Whole pancreas transplantation has had limited clinical success-only 1 of 47 recipients reported to the ACS/NIH Transplant Registry is currently alive. Major problems have been allograft rejection and complications related to either the exocrine pancreas or concomitant transplantation of the duodenum. Experimental approaches to whole pancreas transplantation include use of duct-ligated organs or extraperitoneal transplantation of segmental grafts, with exocrine duct drainage established to either host ureter or a retroperitoneal Roux-en-y loop of bowel. The difficulties with whole pancreas transplantation have stimulated the development of techniques allowing transplantation of islet tissue as a free graft. Transplantation of purified adult islets, of dispersed neonatal pancreas, or of adult pancreatic fragments with exocrine enzymes inactivated can effectively ameliorate experimental diabetes when transplanted to either the intraperitoneal or portal vein sites. In dogs, adult pancreatic fragments containing both islet and exocrine tissue components can effectively ameliorate diabetes when transplanted to the spleen. The first clinical trials of islet transplantation used only small quantities of purified islets. Although insulin requirements of some recipients were reduced, a completely successful islet transplantation in humans has not been accomplished. Procurement of a sufficient quantity of islet tissue and prevention of allograft rejection are major problems, but solutions appear to be forthcoming from the laboratory. The ability of both whole pancreas and islet transplantation to prevent the development or halt the progression of secondary lesions in experimental diabetes provides an impetus to pursue such an approach in humans.