The prognostic value of the type of anaesthesia used for the excision of malignant tumours has been a subject of controversy. Cell-mediated as well as humoral immune responses can be compromised after general anaesthesia, and recurrences may therefore occur more frequently. On the other hand, excision of primary tumours under local anaesthesia might also influence the prognosis unfavourably. The aim of the present study was to determine the prognostic impact of general and local anaesthesia for the primary excision of cutaneous melanoma. Follow-up data of 4329 patients on the Central Melanoma Registry of the German Dermatological Society were analysed. Cox proportional hazards analysis was performed to evaluate the independent significance of the prognostic factors, and survival probabilities were calculated for matched pairs using Kaplan-Meier estimates. Statistical analysis revealed an independent significant effect on survival for tumour thickness, ulceration, level of invasion, anatomical site and gender. General anaesthesia for primary excision of melanoma was associated with a decrease in the survival rate (relative risk 1.46, P<0.0001). This study revealed a slight but significantly increased risk of death for patients treated with general anaesthesia for the primary excision of melanoma. Therefore local anaesthesia should be preferred for the treatment of primary melanoma.