INTRODUCTION: The utility of frozen section performance during partial nephrectomy (PN) is controversial. We assessed the predictive value of frozen sections on final margin status for patients undergoing PN for localized renal tumors. MATERIALS AND METHODS: We queried our prospectively maintained kidney cancer database for patients undergoing PN with localized renal tumors from 2005-2011. Patients were stratified based on the receipt of frozen section analysis into 'frozen' and 'no frozen' groups. Groups were compared using ANOVA, Chi-square, and Wilcoxon's tests. RESULTS: A total of 537 patients (mean age 58.1 years ± 12.0 years, 64.2% male) underwent PN (mean tumor size 3.7 cm ± 2.0 cm; mean Nephrometry score 7.5 ± 1.8) from 2005-2011. Comparing tumor characteristics between patients undergoing frozen sections (83.1%) and those who did not (16.9%), no differences in histology, Fuhrman grade, pathologic stage, or Nephrometry Score were observed between groups. Final margins were positive in 10 patients (11.0%) in the 'no frozen' group compared to 20 patients (4.5%) in the 'frozen' section group (p = 0.01) but in patients with a documented malignancy on final pathology, final margins were positive in 5.5% and 2.9% respectively (p = 0.16). Four patients (0.7%) had local recurrences, all of whom had negative frozen and final pathologic margins. There was no correlation between positive surgical margins and local recurrence (p = 1.0) at a median follow up of 21 months (IQR = 9-31months). CONCLUSIONS: In our institutional cohort, frozen section analysis failed to impact final margin status in patients with documented renal cell carcinoma. Given the oncologic uncertainty of positive surgical margins, further prospective evaluation is necessary to determine the clinical utility of frozen section analysis.