Melanoma vaccines are usually administered after surgical resection of the tumor with the hope of eradicating the micrometastases, in high-risk patients. As we previously reported, most of the melanoma vaccines failed to show any major impact on the disease, except for the autologous whole cell vaccine. This can be explained by the heterogeneous nature of cutaneous melanoma that expresses various levels of melanoma antigens, peptides and has various genetic profiles among different patients. From an immunological point of view, it is illogic to eliminate the tumor and its specific antigens then apply allogenic type of therapy and expect a tumor response. Therefore, it is more logical is to utilize the tumor site as a source for the tumor-specific antigens. In the meantime, patients with in-transit metastases can give us an excellent opportunity to evaluate the local and systemic effects of intralesional (intratumoral) therapy, and various agents have been utilized with equivocal results. On the other hand, intralesional administration of 2 cytokines seemed to process the tumor antigens and activates thymic-derived lymphocytes (T cells). This can induce an antitumor immune response in vivo, i.e., autoimmunization (auto-vaccination), specific to each patient, and overcome tumor heterogeneity regardless to its antigenic or genetic profiles.