Liver metastases of colorectal carcinoma occur in about 50-60% of patients. To improve survival of these patients, there is an urgent need for new treatment strategies. For this purpose, the availability of a preclinical model to develop and test such treatments is mandatory. An ideal animal model for studying liver metastases of colorectal origin should mimic all aspects of the metastatic development in humans and be practical, predictable, and optimal in terms of ethical considerations. Thus far, no model has been developed which satisfies all these conditions. As a consequence, choosing an animal model for the study of liver metastases requires compromises and choices about the necessary characteristics that depend on the purpose of the intended experiments. This overview addresses the advantages and disadvantages of different animal models used for research on experimental liver metastases of colorectal origin. Based on data available in literature, we conclude that heterotopic injection of undifferentiated syngeneic tumor cells in immunocompetent rodents covers most of the desired characteristics. Both subcapsular as well as intraportal injection will yield suitable models and the eventual choice will depend on the aim of the study.