A single prostate specific antigen (PSA) measurement is of limited use in differentiating between prostate cancer and BPH, unless it is significantly elevated. When the most frequently used PSA cut-off point of 4.0 ng/ml is employed, a significant number of patients with BPH are identified as false positives for prostate cancer. Men with a PSA of between 4.0 and 10.0 ng/ml have around a 20 per cent chance of having prostate cancer; in those with a PSA of >10.0 ng/ml, the chance of a positive biopsy is 62 per cent. However, the recent Prostate Cancer Prevention Trial (PCPT) demonstrated that 17 per cent of men who were identified with prostate cancer in fact had a PSA value below 4.0 ng/ml, and some patients had a value below 2.5 ng/ml. Annual measurement of PSA may improve accurate diagnosis, but is also controversial. [With external links to the evidence base]