The ratio of free prostate-specific antigen (f-PSA) to total PSA (t-PSA) in serum, calculated as percent free PSA (f-PSA%), is lower in patients with prostate carcinoma (PCa) than in patients with benign prostate hyperplasia (BPH). This parameter facilitates discrimination between the 2 groups of patients, but there is an overlapping of data. A better understanding of factors influencing this ratio is of practical importance. Therefore, f-PSA% was measured in controls and patients suffering from BPH, PCa and chronic prostatic inflammation with t-PSA concentrations up to 20 microg/l using the IMMULITE assays. The relationships of f-PSA% to clinical situation, age, prostate volume, kind of treatment, and stage and grade of tumor were calculated. Compared with controls or BPH patients, mean f-PSA% values were reduced in PCa patients and in patients with chronic prostatic inflammation. The prostate volume was the most important factor to influence f-PSA%. The difference of f-PSA% between PCa and BPH patients with prostate volumes smaller than 40 cm3 was lost if the prostate volumes exceeded 40 cm3. No relationship of f-PSA% to pTNM stage or grade of tumor was observed. In contrast to t-PSA concentrations, the f-PSA% values were not age-dependent and were not influenced by any kind of treatment in BPH and PCa patients either, which simplifies the use of f-PSA% compared with t-PSA. Thus, for using f-PSA% in clinical practice and for interpreting the data correctly, the advantages shown have to be considered along with the potential limitations of f-PSA%.