Abstract: Regional crop production prediction is a significant component of national food security assessment. Remote sensing has the advantage of acquiring soil surface and crop canopy radiation information, however it is hard to reveal the inherence mechanism of crop growth and yield formation. Crop growth models based on the crop photosynthesis, transpiration, respiration, nutrition are successfully applicable for yield forecasting in simple point scale, however, they are hampered by the deriving of regional crop key input parameters. Data assimilation method which combines crop growth model and remotely sensed data has been proved the most potential approach in regional yield estimation. Hengshui district was taken as the study area. Based on the calibration and regional of WOFOST, the WOFOST model had been used to express the characteristic of time series LAI in crop growth season. To solve the system errors of MODIS-LAI due to the mixed pixels effect, the corrected MODIS-LAI was implemented by combining the field measured LAI data and the MODIS-LAI temporal trend information. Time-series LAI was assimilated through combined corrected MODIS-LAI and WOFOST simulated LAI from green-up to heading stage with EnKF algorithm. The assimilated optimal LAI was used to drive the WOFOST model per-pixel to estimate the regional yield. The results indicated that the precision of yield forecasting was obviously improved with EnKF assimilation, compared with the statistical yield, the coefficient of determination was improved from 0.10 to 0.45 and RMSE was reduced from 2 480 kg/hm2 to 860kg/hm2. The results showed that assimilation of the remotely sensed data into crop growth model with EnKF can provide a reliable approach for prediction regional crop yield and had great potential in agricultural applications. The research can provide an important reference value for the regional crop production estimation.