Clearly, pelvic lymphadenectomy can provide important staging information in the management of prostate cancer, but this benefit is counterbalanced by a modest increase in morbidity and the significant cost of the procedure. It is difficult to provide universal recommendations concerning the indications for pelvic lymphadenectomy. Part of the problem lies in the fact that urologists perform pelvic lymphadenectomy for several different reasons. Some surgeons perform pelvic lymphadenectomy to better counsel patients after radical prostatectomy about their risk for disease progression and for planning adjuvant radiotherapy or hormonal therapy. For these surgeons, preoperative clinical staging parameters do not exclude patients from pelvic lymphadenectomy, and frozen section analysis intraoperatively provides no useful information. Alternatively, the staging information from pelvic lymphadenectomy can be used to justify cancellation of the subsequent prostatectomy should regional spread of prostate cancer be identified, sparing the patient the morbidity of an unnecessary radical prostatectomy. With this approach, despite the false-negative rate of up to 30%, the expense of frozen section analysis seems justified. For this second group of surgeons, the problem becomes balancing the modest morbidity and cost of pelvic lymphadenectomy against the probability that nodal spread of prostate cancer will be missed if the procedure is omitted. The authors consider a greater than 4% risk for missing regional disease to be unacceptable in this setting. Following this assumption, Table 3 outlines parameters for clinical stage, Gleason score, and preoperative PSA within which pelvic lymphadenectomy is indicated. These recommendations are based on [table: see text] predictions from the Partin nomogram, which has been validated using a series of over 4000 patients. For the large number of patients with clinical T1c disease and a preoperative PSA less than 10 ng/mL, bilateral pelvic lymphadenectomy is indicated only if prostate biopsy identifies tumor of Gleason grade 4 or higher. For lower-grade tumors in this patient population, the risk for nodal metastasis was less than 5% in the Johns Hopkins and Mayo Clinic series of over 5800 patients with prostate cancer. For a large pool of patients, the several thousand dollar cost of pelvic lymphadenectomy and the risk for injury to the obturator nerves and vessels, the formation of lymphoceles, and chronic genital edema can be eliminated with low risk. A nomogram-based approach provides only a starting point for a decision analysis framework to determine whether the surgeon should perform lymphadenectomy at the time of radical prostatectomy because current nomograms predict only lymph node positivity. In a decision analysis framework, some patient and physician value is derived from a negative lymphadenectomy. Moreover, the morbidity associated with pelvic lymphadenectomy and the potential inconvenience associated with treating such morbidity also would be factored into the decision. Consequently, a decision analysis framework that takes into account prognostic value, costs, morbidity, and health state uses ultimately will provide the most informative method for determining when pelvic lymphadenectomy is indicated in patients with prostate cancer.