T he purpose of the present study was to evaluate the significance of immunogenetic factors on the survival of pancreatic allografts in beagle dogs. Donors and recipients were leukocyte antigen (DLA)-typed and mixed lymphocyte culture (MLC)-tested. Recipients were made diabetic by total pancreatectomy and immediately implanted intraperitoneally with a vascularized, free-draining (duct unligated) pancreatic segmental (FDPS) allograft. Two groups of dogs were studied. In group I consisting of donor-recipient littermates, recipients were immunosuppressed with prednisone and azathioprine (n = 16 dogs), or not immunosuppressed (n = 4). In group II, recipients were made specifically unresponsive by total body radiation, autologous marrow implantation, and kidney transplantation from DLA-MLC identical donors, 1 yr before FDPS transplantation from the corresponding original kidney donors.Survival of the FDPS grafts in group I was inversely related to pretransplant MLC reactivity, irrespective of DLA genotyped match between donor and recipient. Thus, immunosuppressed high MLC reactors (n = 8) rejected FDPS grafts between 7 and 14 d, whereas immunosuppressed low MLC reactors (n = 8) accepted grafts for 25 to 260+ days, and nonimmunosuppressed low MLC reactors (n = 4) accepted grafts for 9-55 d. Rejection (hyperglycemia) of FDPS grafts was sudden, permanent, and unpredictable despite weekly intravenous glucose tolerance tests with measurements of glucose disappearance rates and serum insulin responses. Nevertheless, serial in vitro cell-mediated lymphocytotoxicity (CML) assays revealed increases in CML before graft rejection in low MLC reactors, and decreases in both CML and MLC responses before graft rejection in high MLC reactors. FDPS graft survival was indefinite (>6 mo) in group II dogs, despite low-grade MLC reactivity (2:4 dogs) and CML responses (4:4 dogs). Biopsies of FDPS grafts at 6 mo in normoglycemic dogs showed disappearance of exocrine tissue and coalescence of islets in both groups I and II, but with less fibrosis in group I (immunosuppressed).These results indicate that (a) pancreatic islets in vascularized grafts (FDPS) may survive indefinitely in the presence of a good tissue match best predicted by MLC testing, (b) tissue specific histocompatibility factors appear to be common enough between kidney and pancreas to allow for long-term survival of both organs transplanted from the same donor, at least in appropriate recipients (group II), and (c) immunosuppression is associated with less fibrosis in FDPS allografts.