Metastatic and chemoresistant melanoma can be a good target of immunotherapy because it is an intractable cancer with a very poor prognosis. Previously, we tested a dendritic cell (DC)-based phase I vaccine, and confirmed that it was safe. In the present study, we performed a phase II trial of a DC vaccine for metastatic melanoma patients with mainly the HLA-A24 genotype, and investigated the efficacy of the vaccine. Twenty-four patients with metastatic melanoma were enrolled into a phase II study of DC-based immunotherapy. The group included 19 HLA-A24-positive (A*2402) patients and 3 HLA-A2-positive (A*0201) patients. The protocol for DC production was similar to that in the phase I trial. Briefly, a cocktail of 5 melanoma-associated synthetic peptides (gp100, tyrosinase, MAGE-A2, MAGE-A3 and MART-1 or MAGE‑A1) restricted to HLA-A2 or A24 and KLH were used for DC pulsing. Finally, DCs were injected subcutaneously (s.c.) into the inguinal region in the dose range of 1-5x107 per shot. The DC ratio (lin-HLA-DR+) of the vaccine was 38.1±13.3% and the frequency of CD83+ DCs was 25.7±20.8%. Other parameters regarding DC processing were not different from phase I. Immune response-related parameters including the ELISPOT assay, DTH reaction to peptide or KLH, DC injection numbers were shown to be related to a good prognosis. The ELISPOT reaction was positive in 75% of the patients vaccinated. The increase of anti-melanoma antigen antibody titer before vaccination was also shown to be a prognosis factor, but that post-vaccination was not. Based on immunohistochemical analysis, CD8 and IL-17 were not involved in the prognosis. Adverse effects of more than grade III were not seen. Overall survival analysis revealed a significant survival prolongation effect in DC-given melanoma patients. These results suggest that peptide cocktail-treated DC vaccines may be a safe and effective therapy against metastatic melanoma in terms of prolongation of overall survival time.