Theoretically ELND may be beneficial for the subset of patients who are likely to have occult metastases at the time of diagnosis. The natural history of melanoma, well described from large data bases, has identified a group of patients with "intermediate-thickness" melanomas as the potential beneficiaries. However, pathologic specimens obtained from ELND have demonstrated a much lower incidence of occult disease (15%) than what would be anticipated based on the estimated risk of developing clinically detectable metastasis, more than 50%. Possible explanations for this discrepancy are as follows: (1) routine pathologic examination fails to detect micrometastases accurately thus, grossly underestimating the true incidence of occult disease and/or (2) patients who are at high risk for metastases have not yet developed regional disease at the time of diagnosis and performance of the ELND. In the latter situation, ELND would provide expectant palliation by preventing regional node relapse but would not be expected to impact on survival. Clearly, however, at least three retrospective studies have demonstrated a survival benefit when using ELND for the treatment of patients in the subgroup with intermediate-thickness melanomas. Prospective studies have not confirmed these same results, but they have not targeted the study to the same group of patients. In order to resolve the differences, we have initiated a prospective randomized multi-institutional surgical trial to ascertain the efficacy of ELND in patients with tumor thickness between 1 and 4 mm. More than 700 patients have been entered on this study. Results should provide valuable insights as to the true incidence of occult regional disease and determine if the removal of this disease translates into a survival benefit. In the meantime, the practice of ELND should be a selective one. The decision should be individualized in an attempt to maximize the potential benefit and minimize the morbidity. Tumor thickness, presence or absence of ulceration, anatomical site of primary, and sex of the patient are all important criteria for estimating the risk of metastasis. The elderly and obese are more likely to suffer from long-term complications of lymphadenectomy, and therefore the risk and benefits must be meticulously evaluated before application of ELND in this patient population is undertaken.