Twenty-five patients with high-risk resected stages IIB, III, and IV melanoma were immunized with a vaccine consisting of the minimal epitope, immunodominant 9-amino acid peptide derived from the MART-1 tumor antigen (AAGIGILTV) complexed with incomplete Freund's adjuvant. The last three patients received the MART-1(27-35) peptide with incomplete Freund's adjuvant mixed with CRL 1005, a block copolymer adjuvant. Patients were immunized with increasing doses of the MART-1(27-35) peptide in a Phase I trial to evaluate the toxicity, tolerability, and immune responses to the vaccine. Immunizations were administered every 3 weeks for a total of four injections, preceded by leukapheresis to obtain peripheral blood mononuclear cells for immune analyses, followed by a post-vaccine leukapheresis 3 weeks after the fourth vaccination. Skin testing with peptide and standard delayed-type hypersensitivity skin test reagents was also performed before and after vaccinations. Local pain and granuloma formation were observed in the majority of patients, as were fevers or lethargy of grade 1 or 2. No vaccine-related grade III/IV toxicity was observed. The vaccine was felt to be well tolerated. Twelve of 25 patients were anergic to skin testing at the initiation of the trial, and 13 of 25 developed a positive skin test response to the MART-1(27-35) peptide. Immune responses were measured by release of IFN-gamma in an ELISA assay by effector cells after multiple restimulations of peripheral blood mononuclear cells in the presence of MART-1(27-35) peptide-pulsed antigen-presenting cells. An ELISPOT assay was also developed to measure more quantitatively the change in numbers of peptide-specific effector cells after vaccination. Ten of 22 patients demonstrated an immune response to peptide-pulsed targets or tumor cells by ELISA assay after vaccination, as did 12 of 20 patients by ELISPOT. Nine of 25 patients have relapsed with a median of 16 months of follow-up, and 3 patients in this high-risk group have died. Immune response by ELISA correlated with prolonged relapse-free survival. These data suggest a significant proportion of patients with resected melanoma mount an antigen-specific immune response against a peptide vaccine and support further development of peptide vaccines for melanoma.