Purpose: Granulocyte/macrophage colony-stimulating factor (GM-CSF) administered locally together with vaccines can augment T-cell responses in animal models. Human experience has been limited to small and uncontrolled trials. Thus, a multicenter randomized phase II trial was done to determine whether local administration of GM-CSF augments immunogenicity of a multipeptide vaccine. It also assessed immunogenicity of administration in one versus two vaccine sites. Experimental Design: One hundred twenty-one eligible patients with resected stage IIB to IV melanoma were vaccinated with 12 MHC class I-restricted melanoma peptides to stimulate CD8 + T cells plus a HLA-DR-restricted tetanus helper peptide to stimulate CD4 + T cells, emulsified in incomplete Freund's adjuvant, with or without 110 μg GM-CSF. Among 119 evaluable patients, T-cell responses were assessed by IFN-γ ELIspot assay and tetramer analysis. Clinical outcomes were recorded. Results: CD8 + T-cell response rates to the 12 MHC class I-restricted melanoma peptides (by day 50) with or without GM-CSF were 34% and 73%, respectively (P < 0.001), by direct ELIspot assay. Tetramer analyses corroborated the functional data. CD4 + T-cell responses to tetanus helper peptide were higher without GM-CSF (95% versus 77%; P = 0.005). There was no significant difference by number of vaccine sites. Three-year overall and disease-free survival estimates (95% confidence interval) were 76% (67-83%) and 52% (43-61%), respectively, with too few events to assess differences by study group. Conclusions: High immune response rates for this multipeptide vaccine were achieved, but CD8 + and CD4 + T-cell responses were lower when administered with GM-CSF. These data challenge the value of local GM-CSF as a vaccine adjuvant in humans. (Clin Cancer Res 2009;15(22):7036-44)