Cancer-related mortality can be decreased by prevention, early detection and improved therapies. Although animal models should be used to evaluate the success of cancer therapies, their usefulness is controversial. Many cancer therapies that have cured tumors in mice have not met with similar success when attempted in humans. Current animal models rely mainly on inoculating cell lines into animals, a method that does not reproduce the natural development of the tumor, both for the kinetics of induction and the anatomical site concerned. In this study, we have used an SV40 T-antigen-transgenic mouse model of prostate cancer in which the tumor spontaneously develops orthotopically with a disease progression that closely resembles the progression of human prostate cancer. We have used this model to test the suitability of adoptive cellular immunotherapy. Transfer of naive cells obtained from a T-antigen-negative congenic animal had significant but partial effects: it prevented development of malignant tumors, leaving just minor foci of residual tumor and/or hyperplasia. Adoptive transfer of memory lymphocytes specific for T-antigen, which is a prostatic self antigen in this model, prevented tumor development and progression without affecting the morphology and function of involved tissues. Treated animals were able to breed, and their survival was greatly increased. These results strongly suggest that adoptive immunotherapy should be successful in treating early stages of human prostate cancer.