A valid experimental model system reflects the system under study and is reproducible. Model systems of prostate cancer that accurately reflect the different disease stages are necessary to ensure a proper experimental design aimed at increasing our understanding of the biology of the disease and such models are essential tools to accelerate development of new therapies for prostate cancer. Until recently, a limited number of experimental systems were available and more suitable models derived from human specimens have only recently been developed and become available for use. In addition, transgenic techniques have also permitted the development of unique mouse models. The difficulty in establishing model systems may reflect the complex requirements necessary for cancer progression and should lead us to interpret results from model systems with caution. It is unlikely that a single model system that faithfully reflects the whole process of cancer development and progression will be developed. However, thoughtful use of the available model systems will permit the study of a significant portion of prostate cancer progression. In this review we summarize the properties of the prostate cancer model systems in use and defined their utility and limitations. This review will guide the investigator seeking models with which to test specific hypotheses pertaining to prostate cancer.