Abstract： AIM:To evaluate the efficacy of enticavir treatment and the model for end-stage liver disease(MELD)as prognostic predictor for the acute-on-chronic hepatitis B liver failure.METHODS:Eighty-four patients with acute-on-chronic hepatitis B liver failure were treated with entecavir 0.5 mg daily and the routine treatment. As a control group, another 99 patients with acute-on-chronic hepatitis B liver failure were treated with the routine treatment for comparison. The MELD scores on the day of their admittance to our hospital and the AMELD scores after 2 weeks of medical treatment were retrospectively analyzed. The mortalities of different MELD score groups were also compared between the treatment group and the control group. Analysis was conducted using two-sample t test, the χ~2 test and the receiver operating characteristic(ROC)curve where appropriate. RESULTS:The area under the ROC curve of MELD score(c-statistic)was 0. 649, while the C-statistic for AMELD score was 0.745. On the first day of their admission, when the MELD score was < 25, the mortality rates in treatment group and control group were 28.2%(11/39)and 57.9%(22/38), respectively(χ~2 =6.928, P=0.008). When it was 25≤ MELD score <30, the mortality rates in treatment group and control group were 42.4% (14/33)and 61.7%(29/47), respectively(χ~2 = 2. 898, P =O. 089). When it was 30 < MELD score < 35, the mortality rates in treatment group and control group were 62.5%(5/8)and 76.9%(10/13), respectively(χ~2 =0.505, P=0.477). If MELD score>35, the mortality rates in treatment group and control group were 50%(2/4)and 100%(1/1), respectively(χ~2 =O. 833, P = 0.600). CONCLUSION:The MELD was prognostic indicators in patients with acute on chronic hepatitis B liver failure with entecavir treatment. The AMELD was better than MELD to be the prognostic predictor. It suggests that entecavir may prevent fatality in patients with acute-on-chronic hepatitis B liver failure if therapy starts early enough when the MELD score is less than 25.