In 35 pancreatectomized dogs, the donor pancreatic tail was transplanted by end-to-side anastomoses between the portal vein and the celiac axis to the femoral vessels of the host. To influence the exocrine pancreas secretion of the graft we applied different methods of drainage of the pancreatic duct: open drainage to the abdominal cavity (Group I, n = 15), occlusion of the duct by Neoprene (Group II, n = 12) and by Ethibloc (Group III, n = 8). Main complications were rejection crisis and pancreatitis. Segmental pancreas allografts were rejected in group I within about 34 days, in group II within about 11 days, in group III within about 13 days. The fasting blood glucose levels in all 3 groups returned to normal 2 days after transplantation. Better results were achieved in the open-duct and Ethibloc-occluded grafts. 3 weeks after transplantation, a delayed and diminished increase of insulin concentration after glucose stimulation was observed compared to the dogs of the control group. Peak insulin values were significantly higher (p 0.01) in group I compared to group II. Histologically acinar atrophy with replacement by fibrous tissue occurred in all groups after 2 weeks. The least severe histological changes were observed in group II. The best results were achieved by the open-duct technique, whereas the occlusion groups showed a high rate of early complications like rejection, venous thrombosis and pancreatitis.