The free form of PSA represents an average of 30% of total PSA. The free PSA level, in relation to total PSA, is expressed as a percentage (of free PSA). The percentage of free PSA appears to be relatively independent of benign prostatic hypertrophy, but it is markedly and significantly decreased in the case of prostate cancer and acute prostatitis. A cut-off value for the percentage of free PSA combined with a cut-off value for total PSA can decrease the number of biopsies indicated for the detection of prostate cancer. This approach avoids 20 to 30% of useless biopsies (specificity), but is accompanied by a 5 to 10% reduction of the number of cancers detected (sensitivity of 90 to 95%) compared to the use of total PSA alone. This 5 to 10% risk of missing a cancer of significant volume in a man under the age of 65 years explains why the use of free PSA to guide prostatic biopsies is not routinely recommended. Despite the promising performance of the percentage of free PSA to improve the indication for prostatic biopsies, the methodology of the studies performed to evaluate this test is not sufficient to validate their conclusions. The percentage of free PSA can be prescribed as a second-line test by the urologist, following a first series of negative biopsies in a man with a high clinical and laboratory suspicion of prostatic cancer, in order to propose a second series of biopsies after three months or to define the frequency of clinical and laboratory surveillance. While waiting for the results of prospective studies in the screening setting, recommendations concerning the use of PSA by general practitioners, who are the first to order this test in the context of screening, can be formulated as follows: The value of free PSA assay for first-line cancer screening has not been validated. A consensus has not been reached concerning the use of free PSA as an indication for biopsies and the frequency of assays. Total PSA assay (cut-off value: 4 ng/ml) remains the reference laboratory test for screening and indication of biopsies. A lower normal cut-off value (between 2 and 4 ng/ml) is currently under evaluation.