Vaccination against Porcine Reproductive and Respiratory Syndrome virus (PRRSv) is widely used to control clinical disease, but the effectiveness appears in some cases to be suboptimal. Field reports have stated the presence of routinely PRRSv-vaccinated but ELISA seronegative sows: the ELISA non-responders. The real extent of this phenomenon (prevalence-origin-consequences) was not yet investigated. In this study, the prevalence of ELISA non-responders was assessed by measuring PRRSv-specific antibodies in 1400 sows, originating from 70 PRRSv-vaccinating sow herds, using IDEXX ELISA (ELISA 1) and CIVTEST E/S ELISA (ELISA 2). Neutralizing antibodies (NAbs) were quantified in a virus neutralization assay. Univariable logistic regression was used to identify herd risk factors for the presence of ELISA non-responders. The global prevalence of non-responders varied from 3.5% (ELISA 1) to 4.1% (ELISA 2), the herd-level prevalence was 40% and the within-herd prevalence ranged from 5% to 20% (ELISA 1) and from 5% to 30% (ELISA 2). The ELISA non-responders had significantly lower NAbs than the ELISA responders. Herds using the combination of one modified live vaccine and one killed vaccine had a significantly reduced risk of having ELISA non-responders. A first assessment of the prevalence and possible consequences of ELISA non-responders has been provided by this study. The clinical importance, origin and underlying immunological mechanisms warrant further research.