Although combined pancreas-kidney transplantation (PKT) has become a valid treatment option for selected type I diabetics, the timing of PKT relative to the degree of nephropathy remains controversial. We analyzed results and morbidity in 30 type I diabetics undergoing PKT after starting dialysis (PKT:D) versus 31 type I diabetics undergoing PKT prior to dialysis (PKT:ND). The two groups were similar with the respect to age, duration and severity of diabetes, gender, race, preservation time, retransplants, sensitization, HLA-matching, and CMV status. The mean preoperative serum creatinine was higher in the PKT:D group (9.9 +/- 3.4 vs. 3.9 +/- 1.9 mg/dl PKT:ND, P < 0.01). All patients were managed with quadruple immunosuppression with OKT3 induction. Actuarial patient survival is 100% (PKT:D) and 96.8% (PKT:ND). Renal and pancreas allograft survival are 97% and 93%, respectively, in both groups. The incidence of rejection, infection, operative complications, reflux pancreatitis, and total hospital days was similar in both groups. Long-term renal and pancreas allograft function and quality of life were like-wise comparable. No adverse coagulation or immunologic effects were noted in the PKT:ND group. Rehabilitation potential favored the PKT:ND group. PKT can be performed safely and effectively in the absence of uremia. In selected type I diabetics with significant nephropathy, we believe that PKT is the best treatment option and need not be considered as preemptive, especially in view of increasing waiting times and the variable progressive nature of diabetic complications.