This study investigates whether capsular overgrowth on alginate-polylysine microencapsulated islets is influenced by (1) the presence of islet tissue, (2) MHC incompatibility between donor and recipient, or (3) the presence of autoimmune diabetes. Encapsulated Albino Oxford (AO, n = 6, isografts) and Lewis (n = 6, allografts) rat islets, and encapsulated human islets (n = 5, xenografts) were implanted intraperitoneally into streptozotocin-diabetic AO rats. Also, encapsulated AO islets were implanted into autoimmune diabetic Bio Breeding/Organon (BB/O) rats (n = 5, allografts). Five isografts, five allografts, and three xenografts in AO recipients and five allografts in BB/O recipients resulted in normoglycemia. Two weeks after implantation, islets containing capsules were retrieved by peritoneal lavage, after which all animals that had become normoglycemic after transplantation returned to a state of hyperglycemia. Recovery rates of the capsules of these successful grafts, expressed as percentages of the initially implanted graft volume, varied from 72% +/- 7% to 80% +/- 9%. The associated pericapsular infiltrates (PCI) were similar in all groups and varied from 3.2% +/- 1.4% to 8.3% +/- 2.6%. Similar recovery rates and PCI were also found with empty capsules. However, the recovery rates of recipients with graft failures were lower and showed more PCI. Immunohistological staining of PCI showed no differences in the types of cells in the PCI on capsules with or without islets. We conclude that this early PCI is a capsule-induced foreign body reaction that is not influenced by MHC incompatibility or by the presence of autoimmune diabetes, and it should be avoided by improving the biocompatibility of the capsules.