Purpose: Routine pathologic examination can miss micrometastatic tumor foci in the lymph nodes of patients with prostate cancer, resulting in confusion during tumor staging and clinical decision-making. The objective of this study was to clarify the significance of micrometastases in pelvic lymph nodes in patients who underwent radical prostatectomy for prostate cancer. Experimental Design:The expression of prostate-specific antigen (PSA) and prostate-specific membrane antigen (PSMA) in 2,215 lymph nodes isolated from 120 patients with clinically localized prostate cancer was assessed by a fully quantitative real-time reverse transcriptase-PCR.We regarded specimens in which either PSA or PSMA mRNAs were positive as proof of the ''presence of micrometastasis.'' Immunohistochemical staining of lymph node specimens with an antibody against PSA was also done. Results: Pathologic examinations detected tumor cells in 29 lymph nodes from 11 patients, and real-time reverse transcriptase-PCR further identified micrometastasis in 143 lymph nodes from 32 patients with no pathologic evidence of lymph node involvement. The presence of micrometastatic cancer cells was confirmed by immunohistochemical staining in 61lymph nodes from 17 patients with pathologically negative lymph nodes.The presence of micrometastases was significantly associated with other conventional prognostic variables, including serum PSA value, pathologic stage, Gleason score, and tumor volume. Biochemical recurrence was detected in 32 patients, 17 of whom were negative for lymph node metastasis by pathologic examination (including 4 patients with pathologically organ-confined disease), but were diagnosed as having micrometastasis. Biochemical recurrence^free survival rate in patients without micrometastasis was significantly higher than in those with micrometastasis irrespective of the presence of pathologically positive nodes. Furthermore, only the presence of micrometastasis was independently associated with biochemical recurrence regardless of other factors examined. Conclusions: These findings suggest that f30% of clinically localized prostate cancers shed cancer cells to the pelvic lymph nodes, and that biochemical recurrence after radical prostatectomy could be explained, at least in part, by micrometastases in pelvic lymph nodes.