INTRODUCTION: PSA serum level measurement in the most important tool in the early diagnosis of prostate cancer patients. However, it is recognised it low specificity is due mainly to prostatic benign diseases. Although it is known that inflammation can contribute on this lack of specificity, there is disagreement in the effect of no symptomatic prostatic inflammatory focus on total PSA and percent free PSA serum levels. AIM: To analyse the biological variability in total PSA and percent free PSA serum levels in patients with biochemical criteria of prostatic biopsy and to compare them with the antibiotic induced variability in a previous urinary infections cohort patients. PATIENTS AND METHODS: We analysed 60 patients with previous urinary infections, normal digital rectal examination and PSA between 4 and 20 ng/ml. We measured total PSA and percent free PSA serum levels. Thirty were treated with 3 weeks of ofloxacin and following a new marker determination. Sextant ultrasound guided prostatic biopsy was performed in all cases. RESULTS: 45 patients demonstrated BPH (29 with prostatitis) and 15 prostate cancer (T1c). Significant variations were found on total PSA serum levels (6.97 ng/ml vs 5.82 ng/ml, p = 0.001) and percent free PSA (14.73% vs 17.77%, p = 0.01) only in treated patients. These differences were significant in BPH and BPH with prostatitis patients but not in prostate cancer patients. Treated patients trend was to decrease PSA (13 treated patients shown PSA < 4 ng/ml vs 2 control patients) and to increase percent free PSA. The median variation of percent free PSA was higher than total PSA and was not influenced by PSA level or prostatic volume. Taking 25 as cut-off of percent free PSA, 18.3% of prostatic biopsies could be avoided in the first determination and 20% in the second. Adding the total PSA reduction, 56% of prostates biopsies in the treated patients could be avoided. CONCLUSIONS: Biochemical criteria of prostatic biopsy could be modified in patients with previous urinary infections due to higher variations on serum markers than those explained by biological variations. These variations could be induced by the antibiotic treatment. These results suggested that the inflammatory focus could influence on total PSA and percent free PSA serum levels.