OBJECTIVES: The combined use of total prostate-specific antigen (PSA), clinical stage, and Gleason score accurately predicts final pathologic stage for men with clinically localized prostate cancer. Recently, the free/ total PSA ratio has been proposed as an adjunct for early detection of prostate cancer. We examined the association between free/total PSA and pathologic stage. METHODS: In a prospective study, 301 consecutive men with clinically localized prostate cancer (average age 58.8 years, range 45-72) underwent a staging pelvic lymphadenectomy and radical prostatectomy. Total PSA and free PSA were measured from preoperative sera. Pathologic stage was determined as organ-confined (OC, n = 169), capsular penetration (CP+, n = 108), seminal vesicle involvement (SV+, n = 13) and lymph node involvement (LN+, n = 11). RESULTS: Overall, 292/301 (97%) of the free/total PSA values were < 25%, and thus suspicious for prostate cancer. Combination of total PSA, Gleason score, and clinical stage predicted well OC (P = 0.00001) and LN+ (P = 0.023); whereas, replacing total PSA with free/total PSA ratio did not improve the prediction of OC (P = 0.0007) nor LN+ (P = 0.03). CONCLUSIONS: The free/total PSA ratio cutoff point of 25% had high sensitivity for prostate cancer among a group of men with clinically localized disease. The free/total PSA ratio did not significantly improve the prediction of pathologic stage provided by total PSA when used alone or in combination with Gleason score and clinical stage. These preliminary data demonstrate that free/total PSA levels provide no additional information for pathologic stage prediction when combined with Gleason score and clinical stage in men with clinically localized prostate cancer.