Although sentinel lymph-node biopsy is accepted as a reliable method of staging of melanoma, it is not without problems to the pathologist. It has been re-emphasised that aggregates of benign naevus cells are not uncommon. Usually these are easily identified by a combination of their benign cytology and location in the fibrous skeleton of lymph nodes. This case represents a combination of an unusual pseudo-malignant pattern in the primary lesion with unusual morphology of the sentinel lymph node. The latter prompted reassessment of the cutaneous lesion as a benign naevus. Confirmation of the diagnosis as cutaneous melanoma by a positive sentinel-node biopsy was averted only by a careful comparison of unusual features of the putative primary and the sentinel lymph node. This case illustrates the need for a rigorous protocol for pathological assessment of sentinel lymph nodes for melanomas to assure detection of all metastases but also to avoid misdiagnosis and over-treatment. It also supports "benign metastases" as the mechanism underlying at least some melanocytes in regional lymph nodes.