Objectives: To examine the relationship between preoperative prostate-specific antigen (PSA) and pathologic characteristics of the prostate gland and prostate cancer at radical prostatectomy in patients with clinically localized disease in the early (1993 to 1998) and late (1999 to 2004) PSA eras. Methods: From January 1, 1993 to December 31, 2004, 2067 patients aged 40 to 80 years with clinically localized prostate cancer underwent radical prostatectomy without neoadjuvant therapy at the Cleveland Clinic. The correlation among the preoperative PSA level, prostate volume, percentage of Gleason pattern 4/5, surgical Gleason score, and cancer volume was calculated using Pearson's and Spearman's tests for the early (1993 to 1998) and late (1999 to 2004) PSA eras. Logistic regression analyses were performed to identify independent predictors of the percentage of Gleason pattern 4/5 and cancer volume during each era. Results: In both eras, the PSA level correlated positively with the percentage of Gleason pattern 4/5, surgical Gleason score, and prostate volume, with nearly identical r values. The PSA level also correlated with the cancer volume in the late PSA era (the only era for which cancer volume data were available). In the multivariate model, biopsy Gleason score, clinical T stage, and PSA level were independent predictors of percentage of Gleason pattern 4/5 in both eras and of cancer volume in the late PSA era. Conclusions: Even in the late PSA era, the preoperative PSA level has retained its predictive value for the percentage of Gleason pattern 4/5 and cancer volume. The PSA level continues to have prognostic value for men with clinically localized prostate cancer treated by radical prostatectomy. © 2007 Elsevier Inc. All rights reserved.