Neurovascular bundle (NVB) preservation during robot-assisted radical prostatectomy (RARP) directly affects patient functional outcomes. Despite careful surgical planning, many NVB preservation techniques are changed intraoperatively from their preoperative plan. Our objective was to identify risk factors predicting intraoperative change in NVB preservation technique during RARP.Prospective data from 578 RARPs performed by a single surgeon between 2010 and 2017 at a tertiary care center. Side-specific NVB preservation technique was planned preoperatively. Surgical techniques were either complete nerve sparing (CNS), or incomplete nerve sparing (INS). Variables included age, tumor grade, prostate volume, number of lifetime biopsies, history of post-biopsy sepsis, and laterality. Variables were modeled in multivariable logistic regressions as potential predictors of deviation in surgical technique. Functional and oncological outcomes were also assessed.A total of 46.9% of cases underwent some intraoperative change in NVB preservation from their preoperative plan. A total of 37.7% of 880 prostate sides planned for CNS underwent unplanned INS. Older age, Gleason ≥ 3+4, post-biopsy sepsis, prostate volume, and left-sided dissections were significantly associated with unplanned INS. Number of lifetime biopsies was not a predictor of unplanned INS. Patients with an intraoperative change to INS had poorer potency and continence. Study limitations included the retrospective nature of analysis and lack of pathological assessment of NVB preservation.Age, Gleason ≥ 3+4, post-biopsy sepsis, prostate volume, and laterality were significant predictors of unplanned INS during RARP, which should guide patient counseling when discussing risks and functional outcomes. The number of lifetime biopsies did not predict unplanned INS, a valuable finding for patients on active surveillance. Our findings highlight the importance of careful preoperative planning and novel adjuncts such as multiparametric MRI.