PURPOSE Our aim was to determine a more predictive cut-off value for free to total prostate-specific antigen ratio (f/tPSA) to better differentiate prostate cancer (PCa) from benign prostate hyperplasia (BPH) in Iranian patients with serum PSA levels between 4 and 20 ng/mL. MATERIALS AND METHODS This study was performed on 332 men with serum tPSA level of 4 to 20 ng/mL. All patients underwent transrectal ultrasound guided biopsies. Serum levels of tPSA and fPSA were measured by Roche immunoassay Elecsys 2010. Relationship between f/tPSA and cases of PCa was determined. RESULTS Prostate cancer detected in 49 (15%) patients. Incidence of PCa for serum tPSA level < 10 ng/mL and serum tPSA level of 10.1 to 20 ng/ mL was 17 (6.7%) and 32 (39.5%), respectively. Mean f/tPSA value was significantly lower in PCa patients (0.12 +/- 0.01) than in benign histology group (0.16 +/- 0.03). Among patients with serum PSA level of 4 to 10 ng/mL (n = 251), mean f/tPSA in benign histology group (n = 234) was 0.16 +/- 0.08 and in PCa group (n = 17) was 0.13 +/- 0.06 (P < .05). For serum PSA level of 10.1 to 20 ng/mL (n = 81), mean f/tPSA in benign histology group (n = 49) was 0.16 +/- 0.08 and in PCa group (n = 32) was 0.12 +/- 0.05 (P < .05). The cut-off value of 0.12 produced 76% sensitivity and 71% specificity, whereas the cut-off value of 0.14 yielded 83.5% sensitivity and 61% specificity. CONCLUSION Determination of f/tPSA ratio improves differentiation of Pca from BPH. We recommend a cut-off value of 0.14 to be applied to Iranian patients.