The invention discloses a region electric vehicle charge load time and space distribution prediction method, comprising the following steps S1, establishing a road network model based on road networkinformation, establishing a power grid model according to power grid information, and establishing a road network-power grid coupling relationship; S2, through combination of a resident travel database and through introduction of travel chains, fitting first-time travel time and residence time at travel destinations of vehicles according to a probability function; S3, planning vehicle travel pathsto obtain travel distance through adoption of a Dijkstra algorithm, and calculating travel driving time and states of charge; and S4, through combination of battery electric quantity levels, judgingcharge demands, determining electric vehicle charge periods and positions, and carrying out repeated sampling through utilization of a Monte Carlo method, thereby obtaining an electric vehicle chargeload time and space distribution prediction result. According to the method, the charge demands of any time, any place and any vehicle can be obtained, and through combination of a road network and power grid geographical coupling property, influences on aspects such as power grid load, network loss and a voltage after electric vehicles are connected are estimated from a time dimension and a spacedimension.