This study aimed to determine the effect of surgical margin positivity on biochemical recurrence (BCR) in patients with locally advanced prostate cancer (PCa) who underwent radical retropubic prostatectomy (RRP). The medical records of all patients with locally advanced PCa that underwent RRP were retrospectively reviewed. Patient demographics, digital rectal examination findings, prostate biopsy Gleason score, prostate volume, pre- and post-treatment prostate-specific antigen (PSA) levels, definitive pathology Gleason score, surgical margin status, seminal vesicle invasion, perineural invasion, absence or presence of BCR, and the time to BCR were analyzed. The study included 130 patients. The final pathologic examination showed that seven (5.4%) patients had T3a disease and 123 (94.6%) had T3b disease. In all, 93 (71.5%) patients had a positive surgical margin [SM(+)], whereas 37 (28.5%) patients had a negative surgical margin [SM(-)]. Among the seven patients with pT3a disease, four (57.1%) had SM(+), whereas 89 (72.4%) of the 123 patients with pT3b disease had SM(-). BCR occurred in 11.8% (11 of 93) of patients with SM(+) and in 45.9% (17 of 37) of those with SM(-) (p < 0.001). Multivariate logistic regression analysis showed that SM(+) was the only significant predictor of BCR following RRP (relative risk, 0.163; 95% confidence interval (0.062-0.433); p < 0.001). SM(+) in RRP specimens is not always indicative of BCR in patients with locally advanced PCa. RRP should be considered an effective treatment choice for selected patients with locally advanced PCa, despite the associated high SM(+) rate. Copyright (C) 2016, Kaohsiung Medical University. Published by Elsevier Taiwan LLC.