Abstract: This study utilizes grape yield data and corresponding meteorological data from Hebei province from 1981 to 2020.Based on the historical meteorological impact index of crop yield abundance,two methods-maximum probability method and weighted average method-were applied to develop grape yield forecasting models for dynamic yield prediction analysis.Model validation results show that the accuracy rates of yield abundance trend prediction were 77.8% and 84.4% for the maximum probability method and the weighted average method,respectively,while the quantitative forecasting accuracy rates were 83.1% and 90.3%,respectively.Model evaluation results indicated that during 15 yield abundance trend forecasts over 5 years,the maximum probability method had 5 errors,while the weighted average method had 3 errors,with quantitative forecasting accuracy rates of 89.9% and 94.0%,respectively.The weighted average method based on the historical meteorological impact index of crop yield abundance performed better in forecasting grape yield in Hebei province.