Whole exome sequencing studies haverevealed the molecular landscape of metastatic CastrateResistant Prostate Cancer (mCRPC) providingnew information about prognostic and predictive factorsof response to therapies. These studies highlightedpotentially actionable targets leading to the beginingof the biomarker-driven era in prostate cancer.Alterations in androgen receptor (AR), DNA repair genes,PI3K-AKT-MTOR pathway or in genes involved incell cycle are frequently observed in mCRPC patientsand may be relevant in the resistance induced mechanismto approve therapy in this setting. Poly(ADP-ribose)polymerase (PARP) inhibitor in BRCA mutatedpatients, pembrolizumab (inmune checkpoint inhibitors)in mCRPC patients with mismatch repair genedefects and microsatellite instability and ipatasertib(AKT inhibitor) in patients with loss of function inPTEN are examples on how molecular information canbe useful to improve treatment selection. Nonethelessthe heterogeneity of advanced PC, the lack of consensusregarding the optimal biological source of analysisand the optimal time and technique for the analisysare still challenges that need to be defined in the nextfuture. The aim is to review the current literature concerningprognostic and predictive marker of responseto therapies in the mCRPC setting.