The management of regional lymph nodes in the patient with primary cutaneous malignant melanoma remains a contentious issue. Early removal of nodes, which may harbour occult microscopic disease before the development of clinically detectable malignant lymphadenopathy, is thought by some to confer a survival advantage and/or improve locoregional tumour control. While tumour burden at lymphadenectomy has been shown to correlate with prognosis, there are conflicting reports regarding the efficacy of elective lymph node dissection (ELND) compared with a regimen of delayed therapeutic dissection for clinically detectable nodal involvement. Evaluation of arguments for and against ELND, together with consideration of surgical morbidity and the predictability of lymphatic drainage patterns, has allowed a clinical algorithm to emerge which may identify patients who may benefit from prophylactic surgery. A chronological summary of the relevant studies is presented and the overall division of opinion regarding the efficacy of ELND is discussed.