Abstract: Based on the surface and upper air observation data from 1981 to 2018, data of surface meteorological elements in Suining observation station, and ECMWF numerical forecasting products from 2013 to 2018.The least square method and MOS method are used to establish the objective prediction method of cold wave in Suining.The results showed that the objective prediction method of cold wave in Suining obviously improved accuracy rate of ECMWF surface air temperature forecasting, and the method proposed can reduce vacancy forecast rate and omissive judgement rate, and the method performed well In actual forecasting applications, five cold wave weather processes from April 2018 to April 2019 were accurately predicted, and which had a high guiding role in forecasting cold wave and cooling weather process in Suining.