We investigated whether local extension findings on preoperative MRI and excisional pathology are associated with positive surgical margin and biochemical recurrence after robot-assisted radical prostatectomy. We identified 704 of our patients that underwent robot-assisted radical prostatectomy between 2012 and 2020, and extracted the 326 patients who had preoperative MRI scans and a radiologist reading. These patients were classified into groups according to the presence of local extension on MRI and pathological findings: ≤ cT2pT2 (195 cases), ≤ cT2pT3 (55 cases), cT3pT2 (31 cases), and cT3pT3 (45 cases). We compared positive surgical margin and biochemical recurrence between them. Median age was 69years, positive surgical margin rate was 20.2%, and five-year biochemical recurrence rate was 20.3%. Of the 226 patients without local invasion on excisional pathology, those with local extension on MRI (cT3pT2) had relatively higher positive surgical margin rate (29.0% vs. 14.4%, p = 0.05) and significantly higher five-year biochemical recurrence rate (25.8% vs. 9.3%, p = 0.01) than those without local extension on MRI (≤ cT2pT2). Similarly, among the 100 patients with local extension on excisional pathology, those with cT3pT3 had relatively higher positive surgical margin (37.8% vs. 21.8%, p = 0.08) and significantly higher five-year biochemical recurrence (53.3% vs. 29.3%, p = 0.01) than those with ≤ cT2pT3. In multivariate analysis, local extension on MRI was an independent predictor of biochemical recurrence (OR 2.1, 95%CI 1.1-3.9, p = 0.01). Local extension on MRI is a prognostic factor independent of pathological stage. The use of MRI may complement the prognostic value of excisional pathology of prostate cancer.