Melanoma has a somewhat unpredictable behavior, and spontaneous regressions do occasionally occur. Many have surmised that these are the result of immunologic attack by the host. Immunologic treatment has been more successful for melanoma than for most other neoplasms, even with relatively crude therapies, such as bacterial products. With the availability of recombinant cytokines, immunotherapy for melanoma has entered a new era. Interleukin-2 (IL-2), which acts entirely through immunologic mechanisms, has been tested extensively, either alone, in combination with other cytokines, or with adoptive cellular therapy. Alone, it has only modest antitumor activity, even at high doses. Its utility may be greater when combined with immunocompetent cells, especially tumor-sensitized T lymphocytes, in adoptive immunotherapy. On the other hand, nonspecific lymphokine-activated killer (LAK) cells do not appear to add significantly to the efficacy of IL-2 alone. Interferon-alpha (IFN-alpha) [corrected] also has had fairly limited activity in the advanced disease setting, but, on the basis of a recently completed randomized trial, has arguably become the "standard of care" in the adjuvant setting for patients with high-risk melanoma, particularly node-positive patients. A number of regimens combining IL-2, IFN-alpha, and chemotherapeutic agents have yielded striking response rates in small trials and await confirmation in larger studies. With better delineation of the host immune response and definition of relevant tumor antigens, we can look forward to exciting results with combinations of vaccines, cytokines, and adoptive cellular approaches, particularly in patients with micrometastatic disease.