Objective: To determine the application value of three-dimensional reconstruction in preoperative assessment of hilar cholangiocarcinoma. Methods: The clinical data of 34 patients with hilar cholangiocarcinoma undergoing hilar resection in combination with partial liver resection were collected. Of the patients, 15 cases had preoperative three-dimensional surgical planning (observational group), while 19 cases did not (control group). The relevant surgical variables between the two groups were compared. In observational group, the coincidence degree between tumor classifications from three-dimensional reconstruction data and postoperative actual findings, and the difference between two- and three-dimensional images for estimation of liver volume were analyzed. Results: In observational group versus control group, the average intraoperative blood loss (525.4 mL vs. 676.3 mL), operative time (5.0 h vs. 5.8 h) and time of tube removal (8.9 d vs. 11.4 d) were significantly reduced (P=0.014, P=0.020, and P=0.037), but the length of postoperative hospital stay and overall incidence of complications between the two groups had no statistical difference (both P>0.05). In observational group of patients, the coincidence rate of Bismuth-Corlette classification by three-dimensional reconstruction to actual results was 86.7% (13/15); the estimated liver resection volume calculated by the two- and three-dimensional techniques showed no statistical difference (404.9 mL vs. 411.2 mL, P>0.05). Conclusion: Three-dimensional reconstruction can provide a precise preoperative assessment for hilar cholangiocarcinoma, so it has certain guiding significance in surgical treatment of hilar cholangiocarcinoma.