To determine the influences of transrectal ultrasonography, prostate-specific antigen (PSA), and heightened public awarencess of prostate cancer stage at diagnosis, we prospectively evaluated our most recent 173 patients who had a pelvic lymphadenectomy from 1987 to 1991. All patients had clinically localized prostate cancer and underwent bilateral limited pelvic lymph node dissections (N = 173); 19 (10.7%) were found to have nodal metastasis. Pathologic tumor stage and grade information was available for 168 patients who had a simultaneous radical prostatectomy. Clinical T-stage data revealed that only one patient had a T3 lesion. Pathologic T stage showed 7.1% to be T1a (12/168), 4.1% to be T1b (7/168), 13.7% to be T2a (23/168), 34.5% to be T2b (58/168), and 40.5% to be T3 lesions (68/168). Metastatic nodal involvement was not seen in any T1a, T1b, or T2a lesions. A Gleason's score of less than 5 lesions was predictive of no nodal metastasis. The clinical stage was upstaged pathologically in none of the T1a, 16.7% of the clinical T1b, 75% of the T2a, and 73% of the T2b, lesions. With regard to serum PSA, 27% of those patients with a level > 20 ng/ml had nodal metastasis (6/22) in this series. Although an elevated PSA was not predictive of tumor nodal metastasis, no patient with a normal PSA had nodal metastasis. Although the distribution of pathologic T stages is similar to that reported in the literature, our low incidence of nodal metastasis may suggest that prostate cancer is being diagnosed earlier. © 1993 Wiley-Liss, Inc.