Abstract: In this paper, by using all kinds of artificial and automatic soil moisture data, the climatic extremes and timevarying thresholds in different levels were added up and found put, and developing a quality control method for data from automatic soil moisture, whith is applied in Sichuan, and including station information check, data loss check, volumetric water content limit check, climatic extreme check and time consistency check (time-varying check and continuity check) . And the 218 S-files from 21 automatic soil moisture stations in 2014 were checked and tested, the amount of data is about 5 million, the quality control results show that data loss accounted for about 1. 6% of total data, the data beyond climatic extremes accounted for about 9% of total data, the data did not pass time consistency check accounted for about 0. 5% of total data, and the data beyond volumetric water content limit accounted for about 0. 3% of total data. In all the doubtful informations, the proportion of beyond climatic extremes is the largest, so the historical climate extreme of all kinds of soil moisture elements need to be further improved.