Background and Study Aims: Endoscopic retrograde cholangiopancreatography (ERCP) is one of the mainstays in the diagnosis and treatment of hepatobiliary and pancreatic diseases, and is also increasingly used for patients with previous Billroth II gastrectomy. The aim of this study was to review our experience of ERCP in patients with Billroth II gastrectomy, and the complications associated with this procedure. Patients and Methods: The records of 110 patients with Billroth II gastrectomy, treated between January 1993 and December 1997, were received retrospectively. Details noted included indications for ERCP, therapeutic interventions, causes of failure, and complications. Results: A total of 110 patients underwent ERCP; the total number of ERCP attempts was 185. The major indications for ERCP were cholangitis (31 %), common bile duct stones (22 %), and jaundice (15 %). The endoscope was successfully passed up to the papilla in 134 exminations (71 %), and selective cannulation was successful in 122 of these (66 %). There were 63 (34 %) failed ERCP attempts. Causes of failure were: difficulty in entering the afferent loop (n = 19, 10 %), failure to enter the duodenum (n = 23, 12 %), endoscope-related bowel perforation (n = 9, 5 %), and failed cannulation ( = 10, 6 %). The other two failures were caused by desaturation in the patient, and inability to distend the duodenum. The major complication of the procedure was perforation, which occurred in 11 examinations (6 %). Of these perforations, nine occurred in the small bowel while the endoscope was being manipulated through the afferent loop; these patients required laparotomy. Two patients had retroduodenal perforations, one occurring after sphincterotomy and one after cannulation. Both patients were successfully managed conservatively. Three patients suffered bleeding after sphincterotomy (3/185 procedures, 1.6 %), and one patient developed acute pancreatitis. These were managed conservatively. The overall complication rate was 8 %. There were two deaths among the patients with small-bowel perforations, and consequently an overall mortality rate of 1& % (2/185 procedures). Conclusions: Most complications of ERCP in patients with previous Billroth II gastrectomy were caused by bowel perforation while the endoscope was being manipulated through the afferent limb. Such perforations are intraperitoneal and require surgical intervention.