Twenty-four subjects with metastatic melanoma were treated on a randomized Phase Ib trial evaluating an autologous tumor lysate-pulsed dendritic cell (DC) vaccine with or without interleukin (IL)-2. The vaccine consisted of autologous DCs obtained from peripheral blood mononuclear cells (PBMCs) cultured in granulocyte macrophage-colony stimulating factor and IL-4 then pulsed with autologous tumor cell lysate and keyhole limpet hemocyanin (KLH). The primary end points of the trial were safety and immune response to vaccine. Subjects were randomized to vaccine administered every other week times 3, vaccine x 3 followed by low-dose IL-2, or vaccine x 3 followed by high-dose IL-2. Immune response was monitored pretreatment and at 2 and 4 weeks after the third vaccine administration. Disease evaluation was performed at 4 weeks after the third vaccination. Therapy was well tolerated with no local vaccine toxicity greater than grade 1 in any arm. IL-2 toxicity was as expected without additional toxicity from the addition of IL-2 to vaccine. Immune response defined as delayed-type hypersensitivity, PBMC interferon-gamma enzyme-linked immunosorbent spot, and PBMC proliferation, to both autologous tumor and KLH were detected in all arms. Interferon-gamma enzyme-linked immunosorbent spot response to KLH (7 of 10 patients) and autologous tumor (4 of 10 patients) were also detected in subjects with available vaccine draining lymph node cells. There were no differences in immune response between treatment arms. No clinical responses were seen. Autologous tumor lysate-pulsed DC vaccine with or without IL-2 was well tolerated and immunogenic but failed to induce clinical response in patients with advanced melanoma.