Aims of the study: Retrospective studies investigating the use of percent free-PSA for early detection of prostate cancer were limited for various reasons: by their use of long-term stored sera, poor mix of non-cancer to cancer cases and the use of only men with PSA values between 4.0 and 10.0 ng/mL. This prospective study investigates the clinical utility of percent free-PSA and complexed-PSA for early detection of prostate cancer in 219 consecutive men presenting for prostate biopsy. Methods: Of 246 consecutive men who underwent ultrasound guided sextant biopsy of the prostate for PSA elevation and/or suspicious digital rectal exam, 219 men had serum total PSA levels between 2.0 and 20.0 ng/mL and were included in this study. Serum total, free and complexed (PSA-ACT) were measured (Hybritech Inc.). Results: Pathologic examinations demonstrated that 72% and 28% of the biopsies were non-cancer and cancer respectively. The mean percent free-PSA was statistically different between the groups (cancer 14%+/-6.4 and non-cancer 18+/-9%, P<0.001) and improved cancer detection. PSA-ACT provided only modest improvement in cancer detection over that of total PSA. Among this cohort of men, the optimal total PSA reflex range for percent free-PSA was 3.0-7.0 ng/mL (38% specificity) with a percent free-PSA cut-off of 20% (95% sensitivity) yet only affected 56% of the cases. Conclusions: PSA-ACT added very little additional value to the clinical utility of total PSA for early detection. Percent free-PSA performed well for all reflex ranges. A sensitivity and specificity of 95% and 20% respectively were obtained using a single cut-off of 25% for percent free-PSA for the group of men with total PSA values between 4.0 and 10.0 and correlated well with recently reported prospective analyses.