Portal-enteric (PE) transplantation of the pancreas allograft provides maintained physiologic drainage, and theoretically the portal delivery of transplantation antigens may have beneficial effects on the graft acceptance leading to improved graft survival. To determine whether the technique of pancreas placement affects the incidence of acute rejection we reviewed our experience in technically successful PE and systemic-bladder (SB) drained simultaneous pancreas and kidney (SPK) transplants performed between 1989 and 1994. Forty-seven recipients were included (SB = 30, PE = 17). All patients received cyclosporine based quadruple immunosuppression and survived at least 1 month. The two groups were comparable in HLA mismatches, cold ischemia time and level of immunosuppression at time of rejection. In the SB group the incidence of rejection was 1.04 kidney rejection/patient and 0.90 pancreas rejection/patient whereas the PE group experienced 0.53 kidney rejection/patient and 0.47 pancreas rejection/patient. The two groups were compared using incidence density statistics due to great variation in follow-up time. The SB group had a significant higher density of both kidney and pancreas rejections (p < or = 0.037 for kidney rejection and 0.058 for pancreas rejection). In addition, while 6 of 30 (20%) pancreas grafts and 4 of 30 (13%) kidney grafts were lost to irreversible rejection in the SB group, only 1 of 17 (6%) pancreas graft and 1 of 17 (6%) kidney graft were lost in the PE group. These data demonstrate, that the PE placement of pancreas allograft affects the rates of acute rejection and graft loss, and imply that there exist some important immunological advantages when the pancreas graft is drained into the portal circulation.