The initial evaluation of melanoma patients should include an assessment of the regional lymph nodes for metastatic disease. Clinical palpation of the nodes is an important but relatively crude index of metastatic melanoma. In clinically normal nodes, the risk of microscopic metastatic melanoma can be predicted by such prognostic factors as thickness, ulceration, and anatomic location of the melanoma. Other prognostic factors, including the level of invasion and growth pattern (nodular and superficial spreading) provided no additional predictive information after these dominant factors were accounted for. Stage I patients with intermediate thickness melanoma (0.76–4.00 mm) have an improved survival rate with elective Iymphadenectomy, while those with thin (<0.76 mm) and thick (≥4.00 mm) melanomas do not benefit from elective node dissection. Stage II melanoma patients with documented nodal metastases have relatively poor survival rates because of a high risk for distant microscopic metastases. A surgical strategy for deciding' about the timing, the extent, and the treatment goals of regional node dissection in melanoma patients is reviewed.