Abstract: The downscaling prediction of summer precipitation in Heilongjiang province was conducted using the model outputs of the National Centers for Environment Prediction Climate Forecast System version 2 (NCEP_CFSv2) from 1982 to 2017.We first analyzed the relationship between summer precipitation and circulation factors in Heilongjiang province and the prediction of circulation factors in key zones,and then selected the circulation factors that have a good correlation between simulation and reanalysis data and between precipitation observations and reanalysis data as prediction factors.We further examined those factors using an optimal subset regression method,established a downscaling prediction model,and finally conducted the prediction performance validation and independent sample prediction using a cross-check method.The results showed that 69% of model downscaling prediction is consistent with the actual anomaly sign.The 6-year independent sample prediction is correct for 5 years,which is better than current operation forecasting results.The model downscaling technique can well predict the trend of summer precipitation in Heilongjiang province once circulation factors could be predicted accurately.In addition,the model downscaling prediction performs better in La Nina years.