Previously, we described VSV-GP, a modified version of the vesicular stomatitis virus, as a non-neurotoxic oncolytic virus that is effective for the treatment of malignant glioblastoma and ovarian cancer. Here, we evaluate the therapeutic efficacy of VSV-GP for malignant melanoma. All of the human, mouse, and canine melanoma cell lines that were tested, alongside most primary human melanoma cultures, were infected by VSV-GP and efficiently killed. Additionally, we found that VSV-GP prolonged the survival of mice in both a xenograft and a syngeneic mouse model. However, only a few mice survived with long-term tumor remission. When we analyzed the factors that might limit VSV-GP’s efficacy, we found that vector-neutralizing antibodies did not play a role in this context, as even after eight subsequent immunizations and an observation time of 42 weeks, no vector-neutralizing antibodies were induced in VSV-GP immunized mice. In contrast, the type I IFN response might have contributed to the reduced efficacy of the therapy, as both of the cell lines that were used for the mouse models were able to mount a protective IFN response. Nevertheless, early treatment with VSV-GP also reduced the number and size of lung metastases in a syngeneic B16 mouse model. In summary, VSV-GP is a potent candidate for the treatment of malignant melanoma; however, factors limiting the efficacy of the virus need to be further explored.