West Nile virus (WNV) is a vector-borne pathogen of global relevance and is currently the most widely distributed flavivirus of encephalitis worldwide. This virus infects birds, humans, horses, and other mammals, and its transmission cycle occurs in urban and rural areas. Climate conditions have direct and indirect impacts on vector abundance and virus dynamics within the mosquito. The significance of environmental variables as drivers in WNV epidemiology is increasing under the current climate change scenario. In this study, we used a machine learning algorithm to model WNV distributions in South America. Our model evaluated eight environmental variables (type of biome, annual temperature, seasonality of temperature, daytime temperature variation, thermal amplitude, seasonality of precipitation, annual rainfall, and elevation) for their contribution to the occurrence of WNV since its introduction in South America (2004). Our results showed that environmental variables can directly alter the occurrence of WNV, with lower precipitation and higher temperatures associated with increased virus incidence. High-risk areas may be modified in the coming years, becoming more evident with high greenhouse gas emission levels. Countries such as Bolivia and Paraguay will be greatly affected, drastically changing their current WNV distribution. Several Brazilian areas will also increase the likelihood of presenting WNV, mainly in the Northeast and Midwest regions and the Pantanal biome. The Galapagos Islands will also probably increase their geographic range suitable for WNV occurrence. It is necessary to develop preventive policies to minimize potential WNV infection in humans and enhance active epidemiological surveillance in birds, humans, and other mammals before it becomes a more significant public health problem in South America.