Sentinel node (SN) biopsy in melanoma patients is an accurate and minimally invasive method of staging and determining prognosis in patients with early-stage disease, and of identifying those patients who may benefit from complete regional lymph node dissection. Careful identification, removal and pathological assessment of SNs is critical to the accuracy of the technique and deficiencies in any of these steps may result in a false-negative biopsy. Pathologists should examine multiple sections from each SN, stained routinely with haematoxylin-eosin and immunohistochemically for melanoma-associated antigens. However, the most appropriate staining and sectioning protocol is not clear from the available evidence. While it appears that more extensive sectioning protocols detect more melanoma, failure rates and false-negative rates of the procedure do not appear to be significantly worse than with less extensive sampling protocols. Micromorphometrical parameters of melanoma deposits in SNs (such as their size, maximal tumour penetrative depth, microanatomical location and percentage nodal cross-sectional area) have been shown to be predictive of regional non-SN involvement and of clinical outcome. Classifying and measuring these parameters can be difficult, depending on the nature and distribution of the tumour deposits in the SN. Because the positive SNs have been removed in all patients included in studies assessing the clinical significance of tiny SN melanoma micrometastases, the likely therapeutic effect of the SN biopsy itself and/or any complete lymph node dissection, as well as lead time bias, are important confounding factors that must be considered when interpreting these studies. Limited data from some retrospective studies with relatively short follow-up suggest that some melanoma metastases may not be clinically relevant, but other studies imply a likely clinical significance of even very small SN metastases. Until there is unequivocal evidence from prospective randomised clinical trials with long term follow-up for the prognostic significance (or lack thereof) of very small SN tumour deposits, it is probably prudent to treat patients with such deposits as SN-positive.