BACKGROUND: The development of lymphatic mapping techniques has facilitated the identification of the sentinel lymph node (SLN), the first node in the regional basin into which cutaneous lymphatics flow from a particular skin area. Previous studies have shown that SLN histology reflects the histology of the entire basin, because melanoma metastases progress in an orderly fashion, involving the SLN before higher nodes in the basin become involved with metastatic disease. It is uncertain whether these orderly cutaneous lymphatic flow patterns are maintained in grossly involved basins. Lymphatic mapping was performed in a population of melanoma patients with clinically palpable lymphadenopathy to address this question. We aimed to determine whether the presence of gross nodal disease in the basin alters lymphatic flow into that basin so that lymphatic mapping techniques are not applicable, and, in patients referred with a grossly involved basin, whether preoperative lymphoscintigraphy should be performed to identify other regional basins at risk for metastases. METHODS: Eight patients presented with grossly palpable disease in the regional basin and underwent preoperative lymphoscintigraphy. All patients with palpable disease and all basins indicated by lymphoscintigraphy to be at risk were dissected. Three patients presented with clinically palpable nodes at the time of diagnosis, and five developed nodal disease on clinical follow-up after undergoing initial wide local excision only. A total of 10 basins in the eight patients were dissected. Of these, eight of the basins had grossly palpable regional nodal disease, and the other two basins were identified by preoperative lymphoscintigraphy as being at risk for metastases. The SLN was identified with intraoperative mapping, harvested, and submitted to pathology. Complete therapeutic lymph node dissections were performed following the SLN harvest in the basins with grossly palpable disease. SLN biopsy alone was performed in the two basins that did not have clinically palpable adenopathy but showed cutaneous lymphatic flow from the scintigram. RESULTS: Sixteen SLNs were harvested from these eight basins with grossly palpable disease, and 14 (87.5%) contained tumor. In each case, one of the SLNs was the grossly palpable node, and in six of the basins (75%) it was the only site of melanoma metastases. An additional 190 higher level, non-SLNs were removed, 32 (16.8%) of which contained microscopic foci of metastatic melanoma (P = .015). The null hypothesis that melanoma nodal metastasis is a random event is rejected. Two patients with trunk melanoma primary sites were identified to have other basins at risk for metastatic disease on lymphoscintigraphy. SLN biopsies were performed in these two patients, and one had microscopic nodal disease in the SLN. CONCLUSIONS: These data support the fact that cutaneous lymphatic drainage patterns are maintained in patients with grossly involved basins, thus buttressing the idea that the SLN is the node most likely to develop metastatic disease. Gross disease in the basin does not significantly alter cutaneous lymphatic flow into the regional basin, as the sentinel lymph node identified under these circumstances is the same as with the grossly involved node. Preoperative lymphoscintigraphy in patients who present with grossly involved nodes in one basin may identify other regional basins with micrometastatic disease and deserves further study in this setting.