The peritoneal cavity is of renewed interest for pancreatic islet transplantation, since it is the preferable site for transplantation of immunoisolated islets. In this study we investigated the minimum islet graft volume needed to restore normoglycemia after free intraperitoneal isogenic transplantation in streptozotocin diabetic rats. Furthermore, graft function was tested by measuring glucose and insulin response to an intravenous glucose load and spontaneously ingested carbohydrate-rich meal. Three graft volumes were used: 8.0-10.0 (group A); 4.0-5.0 (group B); and 2.0-2.3 microliters (group C); 1 microliter contained about 300 islets. All 10 rats in group A and 7 out of 9 rats in group B became normoglycemic for at least 6 months posttransplant, with blood glucose levels not significantly different from normal control animals. Only 3 out of 9 animals in group C became normoglycemic and never for longer than 3 months. The insulin responses to IVGTT in group A and group B were proportional to the grafted islet volume and always significantly lower than those of normal control rats. The insulin response to the test meal showed a similar tendency, which was found to be associated with the absence of preabsorptive insulin secretion. Maximum postprandial blood glucose levels in group A and group B were 0.8 and 1.5 mM higher than in normal control rats. We conclude that intraperitoneal transplantation of at least 4.0-5.0-microliters islet tissue is needed to reverse blood glucose in streptozotocin diabetic rats, and that glucose and insulin levels on IVGTT and test meal in rats with islet grafts of 8.0-10.0 microliters are not completely normalized. It is suggested that the impaired glucose tolerance is due to an insufficient beta-cell mass and a lack of parasympathetic innervation of the transplanted islet tissue.