Malignant melanoma has been shown to be susceptible to T cell-mediated immunity and, therefore, is a candidate for vaccination approaches. Clinical trials using dendritic cells (DC) loaded with peptides corresponding to tumor antigens are ongoing in several institutions, and some promising results have already been published. However, every single peptide-based vaccine can only be used in a patient with a given single HLA type, and this strategy is not appropriate for patients with rare HLA types or with tumors without defined antigens. A clinical pilot study in patients with disseminated melanoma refractory to standard therapy was initiated using a different approach. The authors generated autologous monocyte-derived DC and fused these DC with gamma-irradiated primary autologous tumor cells by incubation in polyethylene glycol. In previous experiments, the authors had shown that these fused cell products are potent inducers of a T-cell response in a mixed lymphocyte tumor cell culture. Seventeen patients were immunized with the cell product by s.c. injection in monthly intervals without any serious side effects. Of these patients, one had a partial response with decrease in size of all evaluable tumor manifestations. In one patient, some of the metastases were regressing despite an overall progressive disease, and one patient achieved disease stabilization for six months. In the responding patient, in parallel to tumor regression, circumscript hair depigmentation occurred. These data show, that a hybrid vaccine of DC and tumor cells can be safely applied and can induce tumor regressions, however, the clinical efficacy of the approach in its present form is insufficient.