Over time, the parameters commonly used to predict pathological stage in men with localized prostate cancer have changed, and there is now little stratification in pretreatment prostate-specific antigen (PSA) concentrations, clinical stages, and biopsy Gleason scores. This prospective study evaluated the utility of complexed PSA (cPSA ) for predicting organ-confined disease in a contemporary series of subjects. The age range of the 420 men was 39 to 72 years (58.2 +/- 6 years). Specimens were collected before radical prostatectomy, and total and free PSA (Hybritech Tandem assays, Beckman Access; Beckman Coulter, Inc., Brea, CA) and total and cPSA (Bayer Immuno 1; Bayer Corporation, Tarrytown, NY) were measured. Pathologic stage was determined from the prostatectomy specimen. Of the 420 men, 316 (75%) had organ-confined disease, and 104 (25%) had non-organ-confined disease (20.7% had extraprostatic extension, 2.6% had seminal vesicle involvement, and 1.4% had positive lymph nodes). Prebiopsy Gleason score distribution was as follows: organ-confined organ-confined, 6 (87%) and 7 (10%); non-organ-confined, 6 (66%) and 7 (30%). Of patients with organ-confined disease, 75% had clinical stage T1c disease compared with 56% for non-organ-confined disease. Using univariate logistic regression, the following variables predicted organ-confined disease: biopsy Gleason score, clinical stage, total PSA, percent free PSA, cPSA, percent cPSA (P <0.05). A multivariate model with biopsy Gleason score, clinical stage, and cPSA had a receiver operator characteristic area under the curve of 0.69. Replacing cPSA with total PSA in this model provided similar information. cPSA and total PSA were highly correlated (r = 0.985). In summary, cPSA was equivalent to total PSA in predicting organ-confined disease. Present and future models and nomograms using PSA as an indicator of pathological stage could consider use of cPSA.