Percent free prostate-specific antigen (PSA) is a promising tool for prostate cancer (CaP) diagnosis. However, its diagnostic performances have not yet been established. The present study was carried out with the aim of evaluating percent free PSA in the most favourable analytical conditions.Eighty-eight patients affected by newly diagnosed, untreated, primary CaP, and 169 cases with biopsy-confirmed, untreated, benign prostatic hypertrophy (BPH) were prospectively enrolled. Abbott AxSYM total and free PSA were measured by the same technician using the same instrument and the same reagent batch.Percent free PSA was more effective than total PSA in differential diagnosis between CaP and BPH in every evaluated dose range of total PSA. In cases with total PSA4 microg/l, percent free PSA could have reduced by about 50% the rate of unnecessary biopsies with a probably still acceptable 93% cancer detection rate. The likelihood of CaP after the determination of percent free PSA was in fact higher than 50% using cut-off points which provide low sensitivity values (i.e. 58% in men aged 50-59 years).Percent free PSA is superior to total PSA in distinguishing primary CaP from BPH in patients with total PSA between 2 and 30 microg/l and in reducing the rate of unnecessary biopsies in men with total PSA higher than 4 microg/l. However, percent free PSA should be cautiously interpreted in decision making in individual patients since post-test probability is relatively low in men aged 50-70 years.