OBJECTIVE: High serum total prostate-specific antigen (PSA) levels have proven to be predictive of concurrent cancer but the role of prostatic intraepithelial neoplasia (PIN) in the production of total (t) and free (f) PSA is still the subject of research. In this study we wanted to discover whether variations in serum fPSA and tPSA levels are caused by PIN. MATERIAL AND METHODS: We reviewed the medical records of 87 patients: in 32 of them the diagnosis of isolated PIN was made from surgical samples (simple prostatectomy, n = 19; radical cystectomy, n = 13); in 30 patients a diagnosis of benign prostatic hyperplasia (BPH) without PIN or prostatic carcinoma was made after simple prostatectomy (n = 20) or radical cystectomy (n = 10); and in 25 patients a clinically significant prostatic cancer was diagnosed and these patients underwent radical prostatectomy. All patients underwent a standard preoperative evaluation, including serum fPSA and tPSA determinations and PSA density. RESULTS: The frequency of isolated PIN in simple prostatectomy specimens was 6.3%. The mean f/t PSA ratios were 17.66% in the 32 patients with PIN, 19.2% in the 8 patients with low-grade PIN, 17.6% in the 24 patients with high-grade PIN, 24.2% in patients with BPH and 13% in patients who underwent radical prostatectomy. CONCLUSIONS: We believe that to make a definitive diagnosis of isolated PIN without carcinoma, study of the whole prostate gland is necessary, in order to definitively exclude the presence of concurrent neoplastic foci. Our data show that PIN does not contribute to tPSA levels and density; however, it may be responsible for a slight reduction in the f/t PSA ratio, with a significant reduction in cases with high-grade PIN (17.6%) compared to those with BPH (24.2%).