INTRODUCTION When lymphatic metastasis occurs, surgery is the primary treatment modality in melanoma patients. Depending on the tumour stage, patients receive a completion lymph node dissection (CLND) when a positive sentinel node is detected. Patients with clinically evident disease of the regional lymph nodes are recommended to undergo a therapeutic lymph node dissection (TLND). The aim of this study was to assess the morbidity of CLND and TLND and to evaluate the Physiological and Operative Severity Score for the enUmeration of Mortality and morbidity (POSSUM) for preoperative risk adjustment of postoperative morbidity. METHODS The hospital files of 143 patients who underwent CLND and TLND for malignant melanoma were analysed. The POSSUM score was used to predict morbidity rates after surgery for the total patient group as well as separated for CLND and TLND patients. RESULTS The overall complication rate was 28.0% and the mortality rate was 0%. The morbidity rate predicted by POSSUM was 32.9%, the mortality 8.3%. Morbidity in patients undergoing CLND was significantly higher with regard to overall wound complications compared with patients with TLND. In these subgroups, POSSUM failed to predict the rates precisely. CONCLUSIONS The POSSUM score predicted the morbidity of the total patient group accurately but failed to predict the rates in the TLND and CLND subgroups. Patients receiving CLND showed the highest morbidity rates. Preoperative sentinel lymph node biopsy therefore has more influence on postoperative morbidity than the physiological parameters represented in the POSSUM physiological score.