HER2/neu is a compelling cancer vaccine candidate because it is overexpressed on some cancer cells relative to normal tissues, it is known to be immunogenic in both animal models and in humans, and it is already known to be targetable by the antibody component of the immune system in the form of monoclonal antibody therapy with trastuzumab. Vaccines offer the theoretical advantage of being able to elicit T-cell responses in addition to antibody responses. HER2 vaccines have been shown to provide benefit in animal models and to be immunogenic in humans. However, the optimal vaccine formulation is not yet known and the therapeutic efficacy of the vaccines in humans has not yet been evaluated. HER2 vaccine approaches currently being tested include peptide-based, DNA plasmid-based, and protein-based vaccines. Our group has developed and started testing a protein-based vaccine composed of both the extracellular domain of HER2 and the carboxyl terminal autophosphorylation portion of the intracellular domain. The extracellular domain was retained to provide for antibody targeting. The kinase domain of the intracellular domain was excluded because of its high degree of homology to other human kinases. The carboxyl terminal autophosphorylation domain was retained because it is the most unique and possibly most immunogenic portion of the HER2 molecule with the least homology to other members of the HER family. The vaccine, termed dHER2, is immunogenic in mice and primates. In animal models it can elicit CD8 and CD4 T-cell responses as well as antibody responses that suppress the growth of HER2-positive cancer cells in vitro and in vivo. Vaccine trials are contemplated in patients with breast cancer that will determine whether the vaccine construct is similarly immunogenic in humans.