Head or neck location of primary cutaneous melanomas has been described as an adverse prognostic factor, but this has to be reassessed after the introduction of sentinel lymph node (SLN) excision (SLNE). Descriptive statistics, Kaplan-Meier estimates and Cox proportional hazard models were used to study retrospectively a population of 2302 consecutive melanoma patients from three German melanoma centres undergoing SLNE. Approximately 10% of the patients (N=237) had a primary melanoma located at the head or neck (HNM). In both the SLN-positive and SLN-negative subpopulation, patients with HNM were significantly older, more frequently men and had thicker primaries compared with patients with tumours in other locations. The proportion of positive SLNs was lower in HNM compared with other locations of the primary (20 vs. 26%, P=0.048). The false-negative rate was higher in HNM (17.5 vs. 8.4%, P=0.05). In patients with HNM, the SLN status was a significant factor for recurrence-free survival but not for overall survival. SLN-negative HNM patients had a significantly worse overall survival than the SLN negatives with primaries at other sites, whereas the prognosis of the SLN-positive patients was similar in both groups. The prevalence of lymph node metastases after SLNE is lower in patients with HNM compared with other melanoma locations. As a result, the prognostic information provided by the SLN for HNM seems less important. Decision making for SLNE in HNM should be carefully balanced considering the potential morbidity of the procedure.