OBJECTIVE To evaluate the results and complications of choledochoduodenostomy and choledochojejunostomy for benign and malignant disease and to review them in the light of the survival of the underlying disorders. METHODS Retrospective analysis of medical records completed by a thorough inquiry for all patients who were lost to follow-up. METHODS Referrals for primary and secondary surgery for obstructive biliary disease to a university hospital from 1974-1987. METHODS After exclusion of patients who underwent a pancreaticoduodenectomy for cancer (Whipple procedure) 113 patients were included in the study (choledochoduodenostomy = CD,N = 64 and choledochojejunostomy = CJ, N = 49). A complete follow-up was achieved in 105 of 113 patients (93%). METHODS An inquiry was made at the civil registration office if the patients were alive or not. The general practitioners of the patients who had died were contacted about the cause of death and the possible biliary symptoms preceding death and the patients who were still alive received a questionnaire which scrutinized all possible complications and side effects of the operation. METHODS Cholangitis, recurrence of the underlying disease or death of the patient. RESULTS Operative mortality was 4.7% following CD and 12.2% following CJ. Procedure-related complications were found in 10.9% and 28.6% respectively. Recurrent cholangitis was not seen after CD and in three patients with a CJ (6.1%). Survival following biliodigestive anastomosis for benign obstruction was comparable for age and sex matched survival. CONCLUSIONS Although CD for choledocholithiasis has largely been replaced by endoscopic papillotomy and although the choice between the two procedures in malignant disease is most frequently dictated by the operative findings, we conclude that the choledochoduodenostomy is a relative simple operation with a low risk of cholangitis.