Background and Objectives: West Nile virus (WNV) is a mosquito-borne virus, member of the Flaviviridae family that can cause severe, sometimes fatal, neurological disorders to horses and humans. With its constant re-emergences in Europe, WNV is a pathogen that threatens the health of these two species, for which there is currently no treatment. Interactions of WNV with its hosts have mostly been studied in mice and cell lines. Although the use of these models have led to major discoveries, they poorly predict the interactions happening between the virus and the human or equine central nervous system (CNS). Recently, three-dimensional (3D) cellular models called human brain organoids (hBOs) have been developed with the aim to better represent the tissue organisation of the human CNS. They are a new alternative to model neurotropic viral infections. However, to our knowledge, equine brain organoids (eBOs) are not yet available. The goals of this study were to develop eBOs and to use them as a pathological model of WNV infection in horses and as a high value support for antivirals identification. Material and Methods: Equine BOs were generated from equine induced pluripotent stem cells and were differentiated for 70 days. Their size and morphology were evaluated by microscopy and cortical markers expression was assessed by RT-qPCR. Equine BOs were infected with 50, 500 or 5000 pfu of WNV-NY99 for 24, 48, 72, 96 hours and 7 days. Three molecules were tested for their anti-WNV activity and increasing doses were added in eBOs culture media 2 hours before and 72 hours post infection. Viral replication in eBOs and antiviral activity of molecules were assessed through quantification of viral RNA (RT-qPCR) and/or viral particles (TCID50) in supernatants. Results: We showed that eBOs are spheroids which grow overtime and express early and late cortical markers. WNV inoculation led to a progressive increase of WNV RNA in the supernatant, showing that the virus replicated efficiently in eBOs. One molecule, out of the three tested, caused a dose-dependent decrease of infectious viral particles in eBOs supernatants. Conclusion: Our results thus demonstrate that eBOs are 3D structures made of neural cells that are permissive to WNV and that they can be used for the identification of antiviral molecules. They represent a novel and unique, physiologically relevant, 3D in vitro model of WNV infection of the equine brain that could accelerate the development of antiviral treatments for horses.