A chimeric human Her2/neu gene (ChHer2) harboring most of the known major histocompatibility complex class I epitopes of the HER2/neu oncogene was expressed as a fusion protein to a non-hemolytic fragment of listeriolysin O (LLO), by the highly attenuated Listeria vector LmddA, which lacks antibiotic selection markers and the ability to spread from cell-to-cell. This construct (ADXS31-164) was tested for immunogenicity and anti-tumor effects in mice. Despite being highly attenuated, ADXS31-164 proved to be efficacious in breaking immune tolerance toward the HER2/neu self-antigen. ADXS31-164 elicited strong T-cell immune responses in experimental animals. In tumors, ADXS31-164 caused a reduction in regulatory T cells (Treg) accompanied by an increase in the CD8(+)/Treg ratio. Comparison of this vaccine with the conventional antibiotic resistant Listeria vector (Lm-LLO-ChHer2) shows that ADXS31-164 is more efficacious in delaying tumor growth in Her2/neu transgenic animals. Because of its well-defined attenuation mechanism and independence from antibiotic selection markers, ADXS31-164 is potentially more suitable for human use. These results support the future clinical development of this vaccine for the treatment of HER2/neu-overexpressing malignancies, such as breast, colorectal and pancreatic cancers.