West-Nile Virus (WNV) is a member of the Flaviviridae family, genus Flavivirus. It is an arbovirus transmitted by mosquitoes which can accidentally infect horses and humans. In both species, it causes neuro-invasive disease that can be lethal. WNV has a worldwide distribution and is currently expanding through Europe. Although vaccines are available for horses, they do not exist for humans and there is currently no antiviral therapies available. To overcome this lack of therapy, it is essential to identify antiviral molecules that will be efficient in the brain. To meet this need, we sought to develop a new in vitro model of equine brain cells. Using induced pluripotent stem cells from horses, we obtained neural progenitor cells that we infected with the NY99 strain of WNV. We used this model to screen molecules for their antiviral capacities by automated cell imaging. Unexpectedly, our results revealed that a family of molecules (called X) exhibited a proviral activity against WNV in these cells. This was confirmed by quantification of viral particles (by TCID50) and viral RNA (RT-qPCR) in the supernatants. We further showed that this proviral effect occurred in a dose-dependent manner. However, this was not observed in other cellular types such as VERO, A549 and equine derm cells (ED). On the contrary, we revealed an antiviral effect of the same molecules on human neural progenitor cells (hNPCs). Our results thus show an opposite role of this molecule in neural progenitor cells depending whether they are of human or equine origin. This calls for extreme caution when extrapolating results obtain in one species to another one and strongly argue in favour of the development of screens based on primary-like cells that are specific of the organ and species naturally targeted by the virus. We are now studying the molecular mechanisms underlying the pro-viral effect of these molecules on the equine neural progenitor cells infected with West-Nile virus.