Background: With evidence showing over treatment, more patients are choosing active surveillance (AS) for intermediate or lower risk prostate cancer (CaP). Genomic profiling is offered to risk stratify patients to aid in management decision−making. This study reports risk discrepancies between National Comprehensive Cancer Network (NCCN) criteria and OncotypeDx Genomic Prostate Score (GPS) and how this influences decision−making in our CaP population. Methods: An inception cohort study was carried out on 56 patients with NCCN very low to intermediate risk CaP who were candidates for AS and underwent GPS testing on prostate biopsy specimens performed within 6mo of entry. GPS provided a score corresponding to a GPS-based risk stratification. Study endpoints: 1) distribution of GPS risk groups within each NCCN risk category; 2) frequency of change to lower or higher risk based on GPS; 3) effect of GPS on physician recommendations and patient choice on disease management. Results: 52/56 patients had sufficient carcinoma on biopsy for a GPS analysis. GPS reassigned risk in 23% (12/52) of patients, with 10 going from NCCN low risk to GPS very low risk and 2 assigned to a higher GPS risk profile (Table). AS was recommended in 19 patients with GPS very low risk group and 8 patients in the GPS-defined low risk group. Physicians recommended treatment to 7 patients with GPS intermediate risk. Patient choice was congruent with physician recommendation in all cases. No patients chose AS when assigned to a higher risk category. All 10 patients reassigned to a lower risk category chose AS. Conclusions: In this CaP cohort, assessment by GPS changed risk stratification in 23% of patients. Moving to a different risk category changed physician recommendation and patient choice in the corresponding direction (to surveillance or therapy) in all cases. More study and larger sample size are needed to fully assess the effect of GPS on clinical decision making.