Large scale of electric vehicles (EVs) integration will have a significant impact on the operation and planning of distribution network, this paper presents an assessment methods of distribution network’s ability of admitting EVs. Based on scenarios of different types EVs integration, which are simulated by Monte Carlo stochastic methods, this paper describes a mathematical model of EVs load prediction. According to the distribution network topology and the mode of power supply, power supply region is divided. Considering the spatial and temporal characteristics of EV charging load and the power supply region, loads are distributed in a hierarchy pattern. Calculating network’s total load and judging whether the total load is overload. The probabilistic model based on the results of three-phase flow assesses the operation situation of distribution network from voltage excursions and network power loss. The IEEE 123 node radial distribution feeders is taken as an example for flexibility evaluation. This method is a powerful tool for distribution network operation and planning.