Porcine reproductive and respiratory syndrome (PRRS) continues to be a threat to the swine industry. Currently, both modified live virus (MLV) and inactivated PRRS vaccines have been licensed for use in the field. However, these vaccines are not always efficacious in protection against infection from a wide array of heterologous PRRS virus (PRRSV) isolates in the field. Safety is also a concern for the MLV, which might be reverted to a virulent strain and subsequently shedding the reverted viruses. In this study, we developed an epitope-based candidate vaccine using a set of consensus B- and T- cell epitopes derived from PRRSV proteins of nsp9, nsp10, GP4, GP5, and N. Since epitopes expressed alone are moderately immunogenic or insufficient in inducing high level of protective immunity, these epitopes were genetically fused with a strong adjuvant, LT192, which is a detoxified bacterial toxin LT (heat-labile enterotoxin produced by E. coli). The epitope-toxin chimera was subsequently transformed in a swine non-pathogenic E. coli strain to use as a live attenuated vaccine. The potential application of this epitope-toxin chimera in PRRS vaccine development was determined in a pig model. The result demonstrated that specific T-cell immune responses were stimulated after immunization. In comparison to the non-immunized pigs, pigs immunized with the candidate vaccine showed improved protection against virulent PRRSV challenge, with about 50% decrease of pneumonic lung lesions and 10-fold reduction of the viral load in serum and lung tissues at 14 days post challenge. This study establishes a vaccine construct platform and swine model system for peptide-based vaccine development against PRRSV and other swine pathogens. Advantages of our candidate vaccine in comparison to MLV or inactivated vaccines includes: 1) Epitope-based vaccines are safer to use than MLV vaccines, since there is no concern about MLV being reverted to virulent strain and subsequently shedding the viruses; 2) It includes the protective T-cell epitopes from PRRSV nonstructural proteins, which are not available from inactivated (killed) PRRS vaccines; 3) A non-pathogenic E. coli strain carries the epitope-toxin chimera construct to the animal, and the epitope-toxin antigen is designed to be expressed while the E. coli colonizes and replicates on the mucosal surface. Such type of vaccine is cheaper to produce and easy to deliver. In the future, it could be packaged with outside coating/feed materials to be produced as powder or pills and delivered as feed-additives, which would be easy to apply and prevent the traditional labor intensive immunization procedures. In addition, epitopes can be easily modified and new epitopes can be included in the construct based on the field epidemic strains. For more information, please contact Dr. Ying Fang at Department of Diagnostic Medicine and Pathobiology, Kansas State University, Manhattan, Kansas; Phone: 785-532-4452""; E-mail: yfang@vet.k-state.edu.