Current research on electric vehicle (EV) charging load forecasting is mostly focused on a short-term single time-scale, and few consider the differences of electric vehicle charging load in different seasons on a longer time scale. Therefore, a multi-time-scale electric vehicle load forecasting method considering seasonal characteristics is proposed. First, a short-term day-to-day electric vehicle charging load prediction model is established taking into account the influence of seasonal characteristics on the initial battery power, mileage power and air-conditioning power consumption of the EV, and combining time and space distribution rules. Secondly, in order to show the characteristics of multiple time scales from short-term (within a day) to medium-and long-term (years), a modified Bass model that takes into account the influence of multiple factors is built to predict the EV holdings in different years in the future. Combined with short-term EV, the charging load forecasting model can be extended to mid-to-long-term EV charging load forecasting, thereby achieving multi-time-scale EV load forecasting and integrating short-term and mid-to-long-term. Finally, through simulation verification with the temperature information and driving data of Shanghai, the results demonstrate that the proposed model is able to effectively predict the EV development trend over the next few years and the EV charging load under multiple time scales considering seasonal characteristics. This work is supported by the National Natural Science Foundation of China (No. 51977180).