Positive surgical margins are a risk factor for recurrence following radical prostatectomy for prostate cancer. Accordingly, positive margins are an indication for adjuvant radiation therapy. The ability to accurately predict the risk of post-operative margin status based on pre-operative clinical characteristics may direct therapeutic management for clinically localized prostate cancer and potentially help avoid treatment with combined surgery and radiation. Based on a large, modern cohort of patients who underwent definitive surgical management for clinically localized prostate cancer, we have designed a novel predictive algorithm for post-operative margin status which incorporates initial prostate-specific antigen (iPSA) level, percent positive biopsy cores (PPC), Gleason score (GS), clinical tumor staging, and post-operative margin status. Records from 1,023 patients treated for clinically localized prostate cancer with radical prostatectomy were analyzed for age, primary and total GS on biopsy, iPSA, PPC, and clinical T stage. A multivariate analysis was performed via logistic regression modelling to create a predictive algorithm. Mean patient age was 59.0 years old. Regarding D’Amico risk stratification, 35.2% were low-risk, 51.5% were intermediate-risk, and 13.2% were high-risk. Post-operatively, 256 patients had positive margins and 767 patients had negative margins. On multivariate analysis, age and primary Gleason score were not statistically significant predictors for positive margins. Total GS, iPSA, PPC, and clinical T stage were all found to be statistically significant predictors. For total GS, odds ratio (OR) was 1.401 (p-value 1.14 x10-4). For iPSA, OR was 1.044 (p-value 5.66x10-5). For PPC, OR was 1.014 (p-value 6.91x10-5). For clinical stage T2a, OR relative to stage T1 was 1.440 (p-value 5.35x10-2). For clinical stage >T2a, OR relative to stage T1 was 2.349 (p-value 3.64x10-2). The area under the curve (AUC) for this model, corrected by bootstrapping, was calculated to be 0.6935. The AUC for this model indicates acceptable predictive performance in terms of sensitivity and specificity for a large, modern cohort of patients. This model may therefore serve to aid in therapeutic decision-making for patients with clinically localized prostate cancer. It does remain to be externally validated. Future modelling will incorporate data from pre-operative magnetic resonance imaging.