Successful induction of functional tumor-specific T cells by peptide vaccination in animal models has resulted in many clinical trials to test this approach in advanced-stage melanoma patients. In this phase I clinical trial, 11 end-stage melanoma patients were vaccinated intradermally with 3 peptides: MART-1(26-35) E27L (ELAGIGILTV), tyrosinase(368-376) N375Q (YMDGTMSQV), and gp100(209-217) T210M (IMQVPFSV), admixed with tetanus toxoid and granulocyte-monocyte colony stimulating factor. The peptide vaccine was well tolerated at all tested doses, and led to grade 1-2 toxicity only. Although all patients did show a rise in antitetanus IgG titers, in only 3 patients peptide-specific CD8 T-cells were induced. In 2 cases, the response was directed against MART-1(26-35) and consisted of 0.2% and 3.3% of the CD8 population; however, in both instances these cells did not produce interferon-gamma on stimulation with the unmodified peptide. The third patient mounted a small (0.1%) response against gp100. In a fourth patient, a nonfunctional tyrosinase-specific response (0.6%) was found that was present before vaccination, but was not affected in size nor in function by the vaccine. None of the 11 patients responded clinically according to response evaluation criteria in solid tumors criteria. Although this study is a small scale phase I clinical trial, the efficacy that was observed was disappointingly low. In accordance with previously published peptide vaccination studies, these results add to the increasing evidence that peptide vaccination in itself is not potent enough as an effective melanoma immunotherapy in advanced-stage patients.