Background: We evaluated the utility of peroral cholangioscopy (POCS) for distinguishing malignant from benign biliary disease to cover low sensitivity of tissue sampling. Methods: From February 1992 to April 2004, all consecutive patients who underwent POCS to confirm the etiology of biliary disorders were included in this study. Brushing cytology or endobiliary forceps biopsy also was performed. We analyzed the diagnostic accuracy of tissue sampling with or without POCS diagnosis. Results: A total of 97 patients (66 men, 31 women; mean age 64.2 years) were included. The final diagnosis was confirmed by surgical resection in 44, clinical follow-up in 52, and cytologic study of ascitic fluid in one. On the basis of ERCP findings, there were 76 strictures and 21 filling defects. Forceps biopsy was performed in 25 patients, and brush cytology was performed in 68 patients. In the remaining 4 patients (4 filling defects, which were identified as stones by POCS), tissue samplings were not carried out. ERCP/tissue sampling correctly identified 22 of 38 malignant strictures and all 35 benign lesions except in 3 patients with inadequate samples (accuracy, 78.0% ; sensitivity, 57.9% ; specificity, 100% ). The addition of POCS correctly identified all 38 malignant strictures and 33 of 38 benign lesions (accuracy, 93.4% ; sensitivity, 100% ; specificity, 86.8% ). For the 21 filling defects observed by ERCP, POCS correctly diagnosed all 8 malignant diseases and 13 benign lesions. Conclusions: The addition of POCS to tissue sampling improves the diagnostic ability and covers for insufficient sensitivity. POCS is especially useful for diagnosing a filling defect. Background: We evaluated the utility of peroral cholangioscopy (POCS) for distinguishing malignant from benign biliary disease to cover low sensitivity of tissue sampling. Methods: From February 1992 to April 2004, all consecutive patients who underwent POCS to confirm the etiology of biliary disorders were We analyzed the diagnostic accuracy of tissue sampling with or without POCS diagnosis. A total of 97 patients (66 men, 31 women; mean age 64.2 years) were included . The final diagnosis was confirmed by surgical resection in 44, clinical follow-up in 52, and cytologic study of ascitic fluid in one. On the basis of ERCP findings, there were 76 strictures and 21 filling defects. Forceps biopsy was performed in 25 patients, and brush cytology were performed in 68 patients. In the remaining 4 patients (4 filling defects, which were identified as stones by POCS), tissue samplings were not c ERCP / tissue sampling correctly identified 22 of 38 malignant strictures and all 35 benign lesions except in 3 patients with inadequate samples (accuracy, 78.0%; sensitivity, 57.9%; specificity, 100%). The addition of POCS correctly identified all 38 the malignant strictures and 33 of 38 benign lesions (accuracy, 93.4%; sensitivity, 100%; specificity, 86.8%). For the 21 filling defects observed by ERCP, POCS correctly diagnosed all 8 malignant diseases and 13 benign lesions. addition of POCS to tissue sampling improves the diagnostic ability and covers for insufficient sensitivity. POCS is particularly useful for diagnosing a filling defect.