We investigated an optimal cutoff level of free/total PSA ratios (f/t PSA) in predicting prostate cancer in different age groups, focusing on the avoidance of unnecessary prostate biopsies. A total of 4955 men were enrolled into the study. Serum tPSA, fPSA, and f/t PSA ratios were determined for the study population. All males who had suspicious digital rectal examination and tPSA > 4 ng/mL underwent transrectal ultrasonography-guided prostate biopsy. Receiver operating characteristic (ROC) curves for each group were generated by plotting the sensitivity versus 1-specificity for the f/t PSA ratio. The sensitivity, specificity, positive likelihood ratio (PLR), and negative likelihood ratio (NLR) were obtained using various f/t PSA ratio cutoffs for different age groups. There were 657 patients with a PSA level of 4-10 ng/mL. According to sensitivity and specificity f/t% PSA cutoff points were determined to be 10%, 15%, 15%, and 10% in 50-59 years, 60-69 years, > 70 years, and all ages categories, respectively, in patients with initial PSA level of 4-10 ng/mL. f/t PSA ratio had an area under the curve (AUC) value of 0.81 (95% confidence level: 0.80-0.82) for all age groups in detecting prostate cancer. f/t PSA ratio has an AUC value of 0.669 (0.632-0.705) in detecting prostate cancer among patients with a PSA level of 4-10 ng/mL. Ten percent of f/t PSA ratio had the highest specificity with PLR and 30% f/t PSA ratio had the highest sensitivity with lower NLR in the all-age categories. The current study shows that the use of f/t PSA ratio in patients with PSA levels of 4 -10 ng/mL should enhance the specificity of PSA screening and decrease the number of unnecessary biopsies. The age-related changes warrant further investigation in a large, multicentric, and multinational population to improve the clinical use of f/t PSA cutoffs. Copyright (C) 2014, Kaohsiung Medical University. Published by Elsevier Taiwan LLC. All rights reserved.