Background: Active surveillance is now accepted initial management for men who have localized prostate cancer with low risk of disease progression. Many criteria have been used for patient identification, including Gleason score (GS) obtained from prostate biopsy. Because of concerns of sampling error, some have recommended repeated biopsy before committing to active surveillance. However, there is limited information about the risk of missing high grade disease using the current standard biopsy approach. This study seeks to compare GS difference from biopsy and surgery to provide an estimated rate of GS upgrade. Methods: The Surveillance, Epidemiology, and End Results (SEER) program was used to identify men with American Joint Committee on Cancer stage T1-2cN0M0 prostate cancer diagnosed between January 2010 and December 2010. Patients who underwent prostatectomy were selected for further analysis. Based on prostate-specific antigen (PSA) levels and GS, cases were divided into low (PSA <=10 and GS <=6) and intermediate (10<PSA<=20 or GS=7) risk groups. The rates of GS upgrade were reported for each group. Chi-square tests were used to assess differences in categorical variables (e.g. age and race) between groups of GS upgrade and no change/downgrade. Results: A total of 10,282 men were evaluated, with 9.2% (n=942) having low-risk disease, and 90.8% (n=9340) having intermediate-risk disease. Among men with low-risk prostate cancer, 22.3% (n=210) had GS upgrade and 0.8% (n=8) had GS 8 disease. Among men with intermediate risk disease, 26.2% (n=2446) had GS upgrade and 2.3% (n=214) had GS 8 disease. There was no statistically significant difference in either age or race distribution among men who had GS upgrade versus no change or downgrade at the time of surgery. Conclusions: A substantial number of low- and intermediate-risk prostate cancer patients had GS upgrade at the time of surgery, but few had upgraded to GS 8 high risk disease. These observations suggest that repeat biopsy prior to active surveillance may not be necessary.