Objective To examine the change in the free/total prostate specific antigen ratio (f/tPSA) with time and to assess the potential value Of serial measurements of f/tPSA as a determinant of disease progression in untreated, low-to-intermediate grade prostate cancer (T1b-T2b N0M0, Gleason score less than or equal to 7 and PSA less than or equal to 15 ng/mL). Patients and methods In a prospective single-arm cohort study from November 1995, patients were conservatively managed with watchful observation alone unless they met arbitrarily defined criteria (clinical, histological and biochemical) of disease progression. Patients were followed regularly and underwent blood tests including PSA and MPSA. The initial and mean f/tPSA and the rate of change of f/tPSA with time were evaluated against the rate constant for the PSA doubling time (PSATd). Correlation analyses were used to evaluate,my association between baseline clinical variables and either the rate of change of f/tPSA or initial f/tPSA. Results As of December 206 accrued patients 2000, 161 of it total of had three or more f/tPSA measurements and formed the basis of the study (median age 70 years; median follow-up 2.7 years). The median initial f/tPSA was 0.16; there was a significant negative correlation between this value and the initial total PSA. The mean f/tPSA and rate of change of f/tPSA with time were significantly negatively correlated with the rate constant for PSATd. Also, the rate of change of f/tPSA correlated negatively with clinical T stage, but not with other baseline variables, including initial PSA, age and Gleason score. Conclusion The f/tPSA in men with untreated, clinically localized prostate cancer varied widely. The negative correlation between the rate of change of f/tPSA with time and rate constant for PSATd suggests that both might provide valuable information to allow clinicians to develop a strategy for optimizing the timing of therapeutic intervention for those patients choosing watchful observation alone.