Proteomic techniques promise to improve the diagnosis of cholangiocarcinoma (CC) in both tissue and serum as histological diagnosis and existing serum markers exhibit poor sensitivities. We explored the use of surface-enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF MS) to identify potential protein biomarkers of CC. Twenty-two resected CC samples were compared with adjacent noninvolved bile duct tissue. Serum from patients with CC (n = 20) was compared with patients with benign disease (n = 20), and healthy volunteers (n = 25). Samples were analyzed on hydrophobic protein chips via SELDI-TOF MS, and classification models were developed using logistic regression and cross-validation analysis. Univariate analysis revealed 14 individual peaks differentially expressed between CC and bile duct tissue, 4 peaks between CC and benign disease, and 12 peaks between CC and sera of healthy volunteers. The 4,462 mass-to-charge serum peak had superior discriminatory ability to carbohydrate antigen 19.9 (CA19.9) and carcinoembryonic antigen (CEA) (P = .004; receiver operating characteristic [ROC] area under the curve [AUC] = 0.76, 0.73, and 0.70, respectively). The training models developed panels of peaks that distinguished CC from bile duct tissue (92.5% sensitivity, 92.3% specificity, ROC AUC = 0.96), CC from benign serum (65.0% sensitivity, 70.0% specificity, ROC AUC = 0.83), and CC from sera of healthy volunteers (75.0% sensitivity, 100% specificity; ROC AUC = 0.92). Serum results were further improved with the inclusion of CA19.9 and CEA (ROC AUC = 0.86 and 0.99 for CC vs benign and healthy volunteer serum, respectively). In conclusion, biomarker panels are capable of distinguishing CC from nonmalignant tissue; serum markers have important diagnostic implications for unknown bile duct stricture.