To assess the value of gadoxetate disodium for characterization and staging of central bile duct stenosis (CBDS).This prospective HIPAA-compliant study was IRB approved. 14 patients (8 male, 6 female; 36–80 years) with clinical suspicion of CBDS underwent preoperative MRI. To estimate the value of hepatocyte phase images (10, 20, 120min p.i.), only T2w images (T2), only post-contrast images (CM), or both image datasets were assessed in three reading sessions by 3 readers. Agreement of each reading session with the intraoperative findings in terms of CBDS etiology and tumor extension (weighted kappa statistic) was calculated.CBDS was caused by hilar cholangiocarcinoma (n=9), gallbladder carcinoma (n=4) and pancreatic carcinoma (n=1). Characterization of CBDS etiology was correct by use of: T2w images in 57%, 64%, 50%; CM images in 64%, 57%, 50%; both in 71%, 64%, 64%. Agreement comparing reading sessions and intraoperative findings regarding tumor extension was fair up to moderate (κ-range=0.21–0.54) as a result of common underestimation. Interobserver agreement for tumor extension was fair (κ-range=0.31–0.33).By means of combined evaluation of T2 and CM images a more reliable characterization of CBDS was possible. Even though CBDS tended to be underestimated assessment of exact tumor extension was improved by contrast administration.