The purpose of this study was to compare the positive surgical margin (PSM) rates of 2 techniques of robot-assisted radical prostatectomy (RARP) for pT2 (localized) prostate cancer. A retrospective analysis was conducted of 361 RARP cases, performed from May 2005 to September 2008 by a single surgeon (KHR) at our institution (Yonsei University College of Medicine). In the conventional technique, the bladder neck was transected first. In the modified ultradissection, the lateral border of the bladder neck was dissected and then the bladder neck was transected while the detrusor muscle of the bladder was well visualized. Perioperative characteristics and outcomes and PSM rates were analyzed retrospectively for pT2 patients (n＝217), focusing on a comparison of those undergoing conventional (n＝113) and modified ultradissection (n＝104) techniques. There was no difference between the conventional and modified ultradissection group in mean age, BMI, PSA, prostate volume, biopsy Gleason score, and D'Amico prognostic criteria distributions. The mean operative time was shorter (p＜0.001) and the estimated blood loss was less (p＜0.01) in the modified ultradissection group. The PSM rate for the bladder neck was significantly reduced by modified ultradissection, from 6.2% to 0% (p＜0.05). In conclusion, modified ultradissection reduces the PSM rate for the bladder neck.