Background: In the era of precision medicine, significant effort has been placed on identifying clinically actionable molecular targets to aid in the treatment of metastatic prostate cancer (mPC). Recent data supports homologous repair and mismatch repair deficiencies to guide the use of PARP inhibitors/platinum chemotherapy or pembrolizumab, respectively. We analyzed the clinical utility of Foundation One (FO) somatic genomic profiling in men with mPC. Methods: We performed a retrospective review of men with mPC in the Duke Cancer Center who received FO testing 01/2010 - 04/2017. We asked whether FO testing identified actionable genomic lesions that led to a change in clinical practice, and if men benefited from this novel genomic matched approach to treatment. Results: We identified 77 men with FO tests. Of these, 77% (59/77) had adequate tissue for FO testing. 76% had mCRPC, 42% had > 3 prior systemic therapies (enzalutamide 51%, docetaxel 39%, abiraterone 41%, sip-T 34%, radium-223 22%). The most common FO genomic alterations were: TP53 deletion/mutation (37%), TMPRSS2-ERG fusion (32%), and PTEN loss (31%). FO revealed 207 genomic alterations, classified into actionable (6), potentially actionable (44), non-actionable (126), and non-informative (31). Of the 6 patients with actionable mutations (BRCA2, ATM, PALB mutation/loss), 4 (67%) received matched olaparib therapy and 3 had PSA or prolonged radiographic stabilization, with durations of therapy of 3-19 months. Of 28 men with potentially actionable genomic findings, 3 (11%) received targeted therapies, including olaparib for a CDK12 mutation, and pembrolizumab for a PMS2 mutation and for a PD-L2 genomic gain. No responses to pembro were observed; the man with a CDK12 mutation responded to olaparib for 8 months. Overall, 42% (25/59) of evaluable men did not have a targetable mutation; 4/77 (5%) had clinical benefits from testing. Conclusions: A small but significant minority (~5%) of men with mPC appear to benefit from FO somatic tumor profiling, particularly those with homologous repair deficiencies. Larger prospective studies are needed with clinical outcomes in order to further understand the clinical utility of routine FO testing in this setting.