Melanoma is a disease which has been shown to be responsive to immune intervention. This has been suggested by reports of spontaneous responses of metastatic disease with strong immune infiltrates, and supported by recent data correlating clinical response after IFNalpha treatment with development of generalized autoimmunity. Since the identification of melanoma-associated tumor antigens, many groups have performed clinical trials to take advantage of this discovery with melanoma-specific cancer vaccines. These trials, in which multiple antigen delivery strategies have been tested in hundreds of patients, have demonstrated that these vaccines are safe, immunogenic, and yield a low frequency of objective clinical responses. The ability to perform careful immunological monitoring has allowed important insights into the nature of the anti-tumor immunity generated by these vaccinations. While many trials have found that the absolute frequency of T cells specific for a vaccine-encoded antigen are a marker of immunization, it does not correlate with objective clinical response. Induction of broad immunity to multiple tumor antigens, taking advantage of cross-reactive T cells and activation of persistent T cells may be more important. Harnessing additional modes of amplifying immune responses (lymphodepletion, cytokine support, inhibition of negative immune self-regulation) are now being tested and should improve clinical responses from 5% to 10% complete response seen currently.