Abstract Objectives. To determine whether lymph node size is a surrogate marker for lymph node metastasis. Methods. We reviewed 980 patients who underwent radical retropubic prostatectomy with bilateral pelvic lymph node dissection for clinically localized prostate cancer, of whom 63 had lymph node metastases. A comparable group of patients with prostate cancer undergoing radical prostatectomy who did not have lymph node involvement was identified using the following parameters: serum prostate-specific antigen level, clinical and pathologic stage, and pre- and postoperative Gleason score. The axial and longitudinal dimensions of the nodes from patients with and without metastases were analyzed to assess the significance of lymph node size in predicting the presence of metastases. All patients had negative preoperative computed tomography (CT) and bone scans. Of the 63 patients with lymph node metastases, 48 had tissue available for measuring the dimensions of the lymph nodes. Results. A total of 76 metastatic and 92 negative lymph nodes were identified from the patients with and without metastatic nodes, respectively. The mean nodal longitudinal size was 1.65 cm (range 0.2 to 6.5) and 3.50 cm (range 0.5 to 9) for positive and negative nodes, respectively ( P = 0.0001). The mean axial nodal size was 0.8 cm (range 0.2 to 3.2) and 1.0 cm (range 0.2 to 2.2) for positive and negative lymph nodes, respectively. In 56 metastatic nodes (74%), the axial size was less than 1 cm and in 20 (26%) less than 5 mm. Conclusions. Lymph node size should not be used as a surrogate for the presence of lymph node metastases. Although no patient had enlarged lymph nodes by CT scan criteria (greater than 1.5 cm), 6 (8%) of 48 and 19 (12%) of 48 patients with and without lymph node metastases, respectively, had nodes with an axial dimension greater than 1.5 cm.