In view of the operational requirements of gridded data at discrete stations and the problems of objective analysis of the Cressman method in areas with complex terrains, based on the 2-m air temperature observed by ground automatic weather stations in and around Shandong, the 2-m air temperature over sea surface predicted by ECMWF , the declining rate of potential temperature by the mesoscale numerical prediction of Shandong, and SRTM elevation data with 90 m resolution, this paper used Cressman method with elevation correction to objectively analyze the 2-m surface air temperature in Shandong, and generated high-resolution ( 0.01°×0.01° ) hourly gridded products of 2-m surface air temperature. The results show the analysis of gridded products using the Cressman method with elevation correction is more reasonable in areas with complex terrains. The monthly mean error is basically within 1 ℃ and the monthly mean error in the mountainous areas of central Shandong is slightly larger than that in the plain areas of the northwest, the southwest, and the southeast of Shandong and Shandong Peninsula. The temperature accuracy in October, November, and December with low temperature is slightly lower than that in May, June, July, August, and September. Each of the test indicators of the products from May to December 2020 performs well, with a mean error of －0.0039 ℃, a mean absolute error of 0.1469 ℃, a root-mean-square error of 0.3597 ℃, an accuracy rate within 2 ℃ of 99.64%, and an accuracy rate within 1 ℃ of 98.24%. In general, the quality of gridded products using the Cressman objective analysis with elevation correction is close to that of HRCLDAS ( High Resolution China Meteorological Administration Land Data Assimilation System ).