In recent years, there have been significant advances in pathological research on alcoholic liver disease (ALD), with suitable animal models making a significant contribution. However, the currently established animal ALD models still have some significant drawbacks, especially the inability to induce the entire human ALD lineage, which may be related to physiological differences between animals and humans. This review comprehensively summarized the most widely used experimental models of ALD, including voluntary drinking, Lieber-DeCarli, Meadows-Cook, Tsukamoto-French, NIAAA, and the "second hit" model. "Second hit" refers to an additional factor that damages the liver. There are various "second hit" models that fall into two main categories: particular diets and drugs. These models can either simulate human drinking patterns more accurately or produce varying degrees of ALD without significantly increasing animal mortality. We introduced the established method of the original models, discussed the advantages and disadvantages of the existing models from the aspects of operability and practicality, and provided existing improvement methods, hoping to provide a reference for future researchers.