In order to establish a model for the in vivo assessment of islet function we have used the Rowett nude rat with transplantation of allogeneic and xenogeneic (mouse) islets into the renal subcapsular space following a minimal period of diabetic induction. Thirty-one nude rats were given streptozotocin and 30 became diabetic with blood glucose levels of greater than 20 mmol/l at 48 h. Rat and mouse islets were prepared by intraductal collagenase and bovine serum albumin density gradient isolation. Eight rats received transplants of freshly prepared allogeneic islets and 8 rats received transplants of 48 h cultured allogeneic islets. Seven rats received transplants of 48 h cultured mouse islets. Diabetes was reversed in all animals and all remained normoglycaemic for 21 days. Graft removal by nephrectomy resulted in hyperglycaemia in 22 out of 23 animals. Histological examination of the grafts showed a band of endocrine tissue beneath the renal capsule which stained strongly positive for insulin and there was no evidence of lymphocytic infiltration/rejection. One rat remained normoglycaemic after graft removal, which may represent recovery of the animal's own islets from the streptozotocin-induced diabetes. Control rats remained diabetic until death. In conclusion, the athymic nude rat can be used for the assessment of allogeneic and xenogeneic islet function when a short (48 h) period of streptozotocin-induced diabetes is used. This model offers a potential method for assessing in vivo function of isolated human islets.