ObjectiveTo establish a predictive model for survival and study it’s clinical value by reviewing the information of patients with hilar cholangiocarcinoma. MethodsMedical record of 196 patients with hilar cholangiocarcinoma were analyzed retrospectively. Seventeen possible clinicopathologic factors were selected. Cox model was used for univariate and multivariate analysis. Prognostic index (PI) was calculated based on the results of multivariate analysis. Patients with different PI were divided into three different risk level groups in order to compare the survival rate. Individual expected survival rate was calculated based on the median PI. Log cumulative hazards function plot was used to test Cox model proportional hazards assumption (PH assumption). ResultsThe significant prognostic factors influencing the survival rate were surgical procedure, surgical margin, and preoperative total bilirubin level (Plt;0.05). The predictive formula was PI=0.815×preoperative total bilirubin level+0.580×surgical margin－0.713×surgical procedure. According to the value of PI, all patients were divided into 3 groups, low risk group (PI≤－0.642), middle risk group (－0.642lt;PIlt;1.364), high risk group (PI≥1.364), and survival rate declined between groups and in groups with statistically significant difference (Plt;0.05). ConclusionThis model for survival can predict the prognosis of patients with hilar cholangiocarcinoma individually and help to conduct individual clinical therapy.