The present study evaluated the therapeutic safety and efficacy of a long-term therapy with a standardised fermented European Viscum album L. (mistletoe) extract Iscador (FME) during post-surgical aftercare of primary intermediate to high-risk malignant melanoma (MM) Union Internationale Contre le Cancer/American Joint Committee on Cancer stage II–III (UICC/AJCC) patients and compared it with an untreated parallel control group from the same cohort. The study was designed as a multicentre, comparative, retrolective, epidemiological cohort study with parallel groups, carried out according to the guidelines of Good Epidemiological Practice. All patients suffered from surgically treated and histopathologically confirmed primary MM in UICC/AJCC stage II–III without distant metastases. In the study group, FME was administered subcutaneously 2–3 times weekly for at least 3 months, while the untreated control group was merely observed (‘watchful waiting’). In both groups some patients also received radio-, chemo- and/or immunotherapy. The patients were followed until the last visit or until death. Unselected, chronologically ordered and standardised anonymous data from medical records that satisfied the predefined eligibility criteria were included for the ‘per protocol’ analysis. Safety was assessed by the number of patients with FME-associated adverse drug reactions (ADRs) and by the search for tumour enhancement. The primary endpoint of efficacy was the adjusted tumour-related survival. Secondary endpoints were overall, disease-free and brain metastasis-free survival. The survival results were analysed after adjustment for baseline imbalances, treatment regimens and other potential confounders by the Cox proportional hazard regression method. Six hundred and eighty-six eligible patients (329 FME vs. 357 controls) from 35 centres were observed for a median aftercare of 81 vs. 52 months. The median FME therapy duration was 30 months. At baseline, both groups were comparable concerning demography, tumour history and risk factors for progression. Additional adjuvant chemotherapy was more frequent in the study group, while immunotherapy was more frequent in the control group. Eleven patients (3.3%) developed systemic ADRs attributed to the FME treatment, and 42 patients (12.8%) developed local ADRs, with mild to intermediate (WHO/Common Toxicity Criteria grade 1–2) ADR severity and spontaneous resolution in most cases. In six patients the ADRs resulted in therapy termination. Life-threatening ADRs, ADR-related mortality or tumour enhancement were not observed. The incidence rate of lung metastases and the adjusted hazard ratio for brain metastases were significantly lower in the FME group. In the course of the study and during aftercare a total of 212 (30.9%) patients relapsed or progressed and 107 (15.6%) died. A significantly longer tumour-related survival was found in the FME group when compared with the untreated controls (unadjusted tumour-related mortality rate 8.9 vs. 10.7%, Kaplan–Meier estimate, log-rank test, P = 0.017), which was confirmed after adjusting for potential confounders by the tumour-related mortality hazard ratio estimate (HR) (95% CI) = 0.41 (0.23–0.71), P = 0.002. The adjusted HR results of overall survival, disease-free survival and brain metastases-free survival were also significantly superior in the FME group.