Permselective acrylic membranes were employed to prevent immune rejection of discordant islet xenografts isolated from various large animals. Canine, porcine, and bovine islets were seeded into tubular diffusion chambers and transplanted into the peritoneum of 27 nonimmunosuppressed streptozotocin-induced diabetic Lewis rats. Six recipients received islet grafts from bovine calves, 7 received grafts from pigs, and 14 received grafts from dogs. Four of the latter were removed at 1 month. In the control group of 10 diabetic rats, 4 received nonencapsulated canine islets, 3 received nonencapsulated bovine islets, and 3 received nonencapsulated porcine islets. Recipients of encapsulated islets promptly dropped from a pretransplantation plasma glucose level of 487 +/- 36 (mean +/- SEM) to 84 +/- 2 (canine), 81 +/- 4 (bovine), and 81 +/- 3 mg/dl (porcine) during the first week. All of the animals sustained these levels for at least 1 month. One rat spontaneously reverted to diabetes at 54 days posttransplantation; 4 other rats became hyperglycemic (glucose, greater than 600 mg/dl) after membrane removal on day 30. The remaining 22 rats maintained fasting euglycemia for greater than 10 weeks. In contrast, rats that received nonencapsulated islets became hyperglycemic in less than 7 days. Intravenous glucose tolerance test K values (decline in glucose levels, %/min) at 1 month for the canine and bovine encapsulated islet transplant group were 3.5 +/- 0.3 and 3.3 +/- 0.1 compared with 3.3 +/- 0.1 (P = 0.63) and 0.91 +/- 0.1 (P less than 0.0001) for normal (n = 4) and diabetic (n = 4) control groups. Morphologic studies of long-term functioning grafts (30-130 days) revealed well-preserved alpha, beta, and delta cells, with varying degrees of granulation. These results demonstrate that immune isolation of islet tissue using permselective artificial membranes can protect discordant islet xenografts from immune rejection in the absence of any immunosuppressive drugs.