Background: Because K-ras oncogene mutations are present in more than 90% of pancreaticobiliary tumors, we tested prospectively whether K-ras mutational analysis of bile samples obtained during ERCP was superior to conventional bile cytology for diagnosis of malignancy. Methods: Bile samples from 20 patients undergoing ERCP for evaluation of jaundice were examined by both cytologic study and the polymerase chain reaction for the presence of K-ras oncogene mutations. Results: Polymerase chain reaction products were amplified from the bile of 8 of 12 patients with malignancy and 3 of 8 with benign disease; K-ras oncogene mutations were present in 4 of 8 polymerase chain reaction products from malignant samples but absent in all 3 from benign samples. No cytologic results were positive for malignancy. Conclusions: Bile obtained during ERCP can yield positive results in K-ras mutational analysis, even when results of conventional bile cytology are negative. In this study, K-ras mutational analysis had a sensitivity of 33% (4 of 12), a specificity of 100%, and a positive predictive value of 100% for diagnosis of malignancy. (Gastrointest Endosc 1995;42:317-20.). © 1995 American Society for Gastrointestinal Endoscopy.