Abstract: Dry-hot wind is one of the agricultural meteorological disasters affecting the yield and quality of wheat. The authors adopted the trend prediction and statistical method to study the occurrence regularity and risk forecast of dry-hot wind, by using the meteorological data and the wheat yield information in Handa . The results showed that: mild dry-hot wind happened twice a year, while severe dry-hot wind happened once a year, and the western mountainous and northern plain regions were the high risk areas for mild dry-hot wind, while the central and eastern plain regions were the high risk areas for severe dry-hot wind; the forecast factors for dry- hot wind risk were first selected through the correlationship analysis between the relative meteorological yield rate and meteorological variables including temperature, humidity and wind speed; the dry-hot wind risk forecast equation was then established using the stepwise regression method. Based on the equation, wind speed and the highest temperature were found to be key factors affecting winter wheat growth during later growth stage.