Objective To select a simple, stable and reliable mouse model of alcoholic liver disease. Methods The mouse models of alcoholic liver disease were induced by oral gavage ethanol or Lierber-DeCarli ethanol liquid diet for 8 weeks. The food intake and body weight were recorded. Pathological changes were examined using HE staining. Liver injury was assessed by the activities of serum ALT, AST, AKP and γ-GT, and serum and hepatic TC and TG. Results After modeling, both models showed significantly increased activities of serum ALT, AST, AKP, and contents of serum and hepatic TG (P<0.05), indicating the successful development of alcoholic steatohepatitis. However, oral ethanol gavage led to body weight loss and weak mental state. Ethanol liquid diet less affected the body weight and mental state. Ethanol liquid diet enhanced liver to-body weight ratio and serum TC, but oral gavage of ethanol did not. The changes of serum ALT, AST, serum and hepatic TG, and hepatic steatosis in the ethanol liquid diet models were more severe than those in the oral gavage ethanol models, suggesting that Lierber-DeCarli ethanol liquid diet led to more serious liver injury than oral gavage ethanol. Conclusions Lierber-DeCarli ethanol liquid diet model is better than oral gavage ethanol model, and is more suitable for studies on mechanisms and evaluation of hepato-protective drugs for alcoholic liver disease.