Many tumor antigens have been identified that can be targeted by the immune system. Animal models that have been genetically modified to express human HLA molecules instead of their own MHC antigens have shown to be valuable in the discovery of peptides derived from tumor antigens many of which have since been used in clinical trials with varying degrees of success. Although these models are not perfect, they nonetheless allow transplantable tumor models to be developed to evaluate novel vaccination strategies that can then be applied in humans. In addition animals that have been genetically modified to "spontaneously" generate tumors that will grow within their correct environment are of greater value for studying angiogenesis, metastasis and the relationship between the immune system and tumor in a physiological setting. In this review, mice genetically modified to express HLA genes or to spontaneously develop tumors are discussed, highlighting their advantages and limitations as preclinical models for cancer immunotherapy.