The objective was to test the hypothesis that in patients with prostate cancer undergoing radical prostatectomy (RP), diabetic patients are at a higher risk of harboring a high-grade tumor than non-diabetic patients. We examined 2060 consecutive men who underwent RP between 2001 and 2009. Of them, 7.1% had type 2 diabetes mellitus (DM). A high-grade tumor was defined as having a Gleason score ≥ 8. Univariable and multivariable logistic regression analyses were used to test the relationship between type 2 DM and high-grade tumor. Mean patient age was 64 years (range: 45-85). Mean total PSA level was 9 ng ml(-1) (range: 1-89.5). A significantly higher percentage of diabetic patients had high-grade tumor on biopsy (16.3 vs 7.6%; P = 0.001) and on RP specimen (21.1 vs 11.7%; P = 0.001) in comparison with non-diabetic patients. In multivariable analyses, DM was an independent predictor of high-grade tumor on biopsy (odds ratio = 2.31, P = 0.001) and on final pathological specimen (odds ratio = 2.22, P = 0.002). In patients undergoing RP, those with type 2 DM had a higher risk of harboring a poorly differentiated tumor on final pathological examination.