The prediction of electric vehicle charging load is the basis of the planning of electric vehicle charging facilities, and it is necessary to analyze the user’s travel law and charging law, and the structure and attributes of the road network are important factors affecting the user’s travel law and charging law. At present, the more common road network structure model is based on the topological structure model of graph theory, which ignores the traffic speed characteristics and bending characteristics of the road, and there is a big difference compared with the actual road, which causes a large error in the load prediction of electric vehicles based on the model. Therefore, an ArcGIS-based electric vehicle charging load prediction method to improve the accuracy of road network topology is proposed. Firstly, OSM (Open Street Map) obtains the road network vector data and imports it into ArcGIS to analyze its road properties. Python is used to process the real-time road condition layer of AutoNavi Open Platform to analyze the congestion of road sections and functionally partition the road network coverage area. Secondly, based on the 2017 National Household Travel Survey (NHTS), the travel behavior of electric vehicle users was analyzed. Finally, the Monte Carlo method is used to simulate the charging behavior of electric vehicle users in the area, and the charging load demand of each functional area is obtained. Taking the urban area of Lanzhou City as an example, the results show that the method can accurately simulate the charging law of users and can intuitively reflect the charging load distribution characteristics of each functional area in the area at different times.