The androgen receptor (AR) is a key oncogenic driver in prostate cancer and androgen deprivation therapy (ADT) is the mainstay of treatment for metastatic disease. Typically there is a very good initial response to ADT, but treatment failure invariably occurs resulting in a disease state referred to as castration-resistant prostate cancer (CRPC). Currently there are no curative treatments for CRPC. Recent studies employing new generation androgen signalling targeting agents such as the androgen biosynthetic inhibitor abiraterone acetate, or the new generation AR antagonist enzalutamide, to more effectively reduce circulating androgen levels or silence the receptor, have revealed diverse adaptive mechanisms to sustain AR signalling. For example, alternatively spliced, constitutively active variants of the AR lacking the ligand-binding domain have been identified that potentially drive AR signalling in an androgen-deplete state and play a key role in progression to CRPC and, indeed, to lethal forms of the disease. As a consequence, on the horizon are new targeting strategies aimed to outsmart known adaptive mechanisms, particularly in terms of targeting regions of the receptor that lie outside of the ligand-binding domain. While these new strategies hold great promise for more effective AR inhibition, historical experience of employing ADT raises a key question: will more effective AR silencing lead to more durable disease regression and significantly extend life for men with metastatic prostate cancer or will it reveal yet unknown treatment-induced adaptations that allow the disease to persist? To assess this, we have developed a unique ex vivo culture approach where human tissues are cultured as explants while maintaining tissue integrity, cell proliferation and androgen signalling. We are utilizing this ex vivo strategy to evaluate multiple new AR-targeting agents, including next generation heat shock protein-90 inhibitors, new androgen biosynthesis inhibitors and novel peptidomimetics that target multiple functional domains of the AR.