A major problem in pancreas transplantation is the lack of a reliable method for the early detection of rejection. Over a 32-month period, we performed 61 combined pancreas-kidney transplants with pancreaticoduodenocystostomy. All patients received quadruple immunosuppression with OKT3 induction. Urine cytologic monitoring was performed on Papanicolaou-stained membrane filters for cell counts and cytocentrifuge preparations for HLA-DR antigen staining. The final diagnosis of rejection was based on clinical criteria, a rise in serum creatinine, histopathology, and hypoamylasuria. Cytologic features of acute rejection included hypercellularity with lymphocyturia, increased numbers of epithelial cells and positive antibody staining for HLA-DR antigen. A total of 36 definite acute rejection episodes occurred in 28 patients, with 19 confirmed by histopathology. Satisfactory urine cytologic specimens were available in 28 rejection episodes and corroborated the diagnosis in 21 cases, for a sensitivity of 75% compared with 75% and 50% with serum creatinine and urine amylase, respectively. When the urine cytologic score was combined with HLA-DR antigen staining, sensitivity improved to 93%. Thirteen false-positive diagnoses occurred in the remaining 1444 urine cytologic specimens available for evaluation, for a specificity for rejection of 99%. The positive predictive value of cell counts was 62% and negative predictive value was 99%. Patient survival is 98.4%, kidney allograft survival is 96.7%, and pancreas allograft survival is 93.4% after a mean follow-up of 15 months. Only 2 immunologic graft losses occurred (1 kidney, 1 pancreas). In conclusion, urine cytologic monitoring shows promise as a simple, reliable, and noninvasive method to detect rejection after combined pancreas-kidney transplantation with bladder drainage. Prospective studies are needed to assess the role of urine cytologic monitoring after solitary pancreas transplantation.