The availability of a DIVA (differentiating infected from vaccinated animals) vaccine is very important for the control and eradication of endemic infectious diseases such as porcine reproductive and respiratory syndrome (PRRS). Previous studies in our laboratory identified several B-cell linear epitopes consistently recognized by convalescent sera obtained from pigs infected with a North American porcine reproductive and respiratory syndrome virus (PRRSV) strain. To ascertain if one or more of these immunodominant epitopes can be used as the basis of DIVA differentiation, we selected two epitope markers previously identified on the non-structural protein 2 (PRRSV NSP2, predictably the viral protein most likely to tolerate large deletions). The choice of these epitopes was primarily based on their immunodominance and their deletion were performed along the backbone of the wild-type cDNA infectious clone (FL12). We were able to successfully rescue a mutant that fulfilled the requirements for a DIVA marker strain, such as: efficient growth of the deletion mutant in vitro and in vivo and induction of specific seroconversion as measured by a commercial ELISA kit, with absence of a marker-specific peptide-ELISA response in 100% (n=15) of the inoculated animals. In summary, our results provide proof of concept that DIVA PRRSV vaccines can potentially be developed by deletion of individual "marker" immunodominant epitopes.