Abstract: Drought and its accurate prediction have always been a major problem in the field of hydrometeorology.To explore accurate and feasible drought forecasting methods，a land-air coupled drought forecasting model based on the VIC model and the European Center for Medium-Range Numerical Weather Forecasting （ECMWF） forecasting meteorological data was established to calculate monitored soil water content and predicted soil water content and to analyze the spatial and temporal distribution characteristics of the Soil Moisture Anomaly Percentage Index （SMAPI），based on which we assessed the drought forecasting ability of the coupled model over Zhejiang Province.The results show that：① for the foresight period，the forecasted and monitored SMAPI values in the foresight period of 0~1l d perform better in terms of absolute error，RMSE，and R-value，which shows the model has the best forecasting effect in the 10-day scale；② The 10-days scale forecast of the model can accurately grasp the development trend of the future droughts in the process of temporal change，and the consistency of the spatial forecast with the grid distribution of the monitored SMAPI values can reach 90% or more；③ The model has better prediction performance in the plain area with little rainfall in the northwest of Zhejiang Province，and better describes the trend of drought prediction in the area with more drastic changes in soil water content.The research results can provide scientific and technological support for drought disaster prevention and early warning in Zhejiang Province.