Background: Pembrolizumab is approved for patients with metastatic, microsatellite instability-high (MSI-H) or mismatch repair-deficient (dMMR) solid tumors who have progressed on prior therapy and have no satisfactory treatment options. However, very few men with PC were included in these initial studies. We evaluated the clinical activity of pembro in a cohort of men with mCRPC and their responses based on molecular genotype. Methods: We performed a retrospective, IRB-approved review of all men with mCRPC in the Duke Cancer Center who were treated with pembro in order to define the duration of therapy, time to PSA-progression, and imaging responses according to prior/concurrent therapy and by molecular subtypes. Results: We identified 51 men who received ≥1 cycle of pembro for mCRPC. Of these, 86% (44/51) had ≥3 prior lines of therapy after ADT, including abiraterone (88%), docetaxel (86%), enzalutamide (enza, 80%), and sipuleucel-T (74%). Somatic tumor sequencing was available in 18/51 men (35%). We found that 16% (8/51) had a ≥50% confirmed PSA decline with pembro, with 8% (4/51) having ≥90% PSA decline. Two of 4 had mutations in LRP1b, one of whom also had MSH2loss and was MSI-H and TMB-high; of the other 2 patients, one had no actionable mutations and the other had no genetic testing available. Fifty-nine percent (30/51) of men were treated with concurrent therapy with pembro (most commonly enza); however, all patients with PSA responses who were concurrently treated with enza had prior documented PSA progression on enza. Overall, the median time to discontinuation of treatment was 5.8 months (mo), median PSA-PFS, defined as a 25% increase in PSA from baseline, was 1.4 mo, and median OS was 6.7 mo. Among patients with ≥50% PSA decline, duration of response ranged from 3.7 to 16.3 mo with 5 patients with ongoing responses. Conclusions: In a heavily pre-treated population of men with mCRPC, pembrolizumab was associated with a ≥50% PSA decline in 16% (8/51) of men, including a dramatic ≥90% PSA response in 8% (4/51), two of whom harbored LRP1b mutations, one also with MSH2loss and a MSI-H tumor. Large prospective studies with genomic testing are needed to identify men with the greatest chance for response to immunotherapy.