Abstract Although androgen deprivation is the standard therapy for metastatic prostate cancer the mechanisms for the initial response and subsequent relapse are largely unknown. The stroma and vasculature are necessary for castration-induced involution of the normal prostate, and androgen-independent tumors respond to castration when surrounded by normal prostate tissue. This suggests that the microenvironment is highly important for the therapeutic effect of castration. To further explore this, androgen-sensitive rat prostate tumor cells (Dunning G) were injected into the rat prostate and/or bone marrow (to simulate bone metastasis) and analyzed for response to castration. Tumors in the prostate responded well to castration with decreased tumor size, both in already established tumors and when tumor cells were injected in castrated animals. The size of bone metastases was, however, unaffected by castration. Both groups developed lung metastasis but castration tended to increase rather than decrease pulmonary tumor burden. To further explore the mechanisms behind this we have established in vitro cell lines from bone metastasis of both castrated and control rats. In our ongoing experiments we are examining if these cells are responsive to castration if instead placed within the prostate environment or if the resistance to therapy has become stable. The response to castration is also further explored in lung metastasis. Our results suggest that the metastatic environment is highly important for the establishment of castration resistant prostate cancer and that the mechanisms behind this need to be further explored. Citation Format: Sofia Halin Bergström, Stina Rudolfsson, Tove Dahl Scherdin, Anders Bergh. The microenvironment decides how prostate tumors respond to castration treatment [abstract]. In: Proceedings of the AACR Special Conference on Advances in Prostate Cancer Research; 2012 Feb 6-9; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2012;72(4 Suppl):Abstract nr B68.