The possibility of maintaining and studying human prostatic cancers in an in vitro and in vivo environment has allowed the development of rare but essential tools to study many aspects of the biology of these cancers. Although none of the available models is perfect, the sum of the studies conducted with these models over more than 20 years constitutes the basis for major progress in our understanding of this disease. The most widely used cell lines (cultured in vitro) are PC-3, DU-145 and LNCaP. They are limited by the fact that they are essentially androgen-independent cell lines, derived from metastatic sites. Due to the slow growth of prostatic cancers, it is very difficult to obtain cancer cell lines which can be transplanted in immunodeficient animals (such as athymic mice) and only a few xenografts are currently available, some of which, like PCV-82, LuCaP 23, CWR-22, are androgen-dependent or sensitive. These models can be used to study host-tumour interactions as well as endocrinological interactions, stroma-tumour cell interactions, and to analyse molecular phenomena related to stages of hormone dependence and hormone resistance. In vitro and in vivo models of metastatic prostatic cancer have also been developed and appear to have a crucial impact on the understanding of metastatic mechanisms and new therapeutic approaches. This paper describes the main experimental models developed from human prostatic cancers, their main characteristics, their value compared to clinical cancers and some of the major studies conducted with these models. Due to the exponential progress in molecular biology techniques and oncogenetics, it appears essential to increase the number and diversity of experimental models of prostatic cancer in order to advance research concerning the crucial phenomena occurring during the course of this disease, from oncogenesis to currently incurable metastatic stages.