In the study of land salinization classification, researchers are most concerned about the distribution, area and degree of salinization. Traditional classification methods of land salinization only manually extract the sample data from the study area, which cannot obtain the classification results for large area. With the development of remote sensing technology, remote sensing data is often used to extract and analyze the information of saline-alkali soil. At present, most classification methods of land salinization utilize the spectral information of remote sensing data based on supervised classification or unsupervised classification, which still have some errors in the classification results. Combining the sample data with Landsat TM images, the Western Jilin Province of China was selected as the study area in this paper. Through analyzing the relationship between the spectral characteristics and the content of soil salinity of the sample data extracted from different types of saline-alkali land, a land salinization classification method using the decision tree was proposed. The experimental results demonstrated that the proposed method can supply more accurate classification information of land salinization, and further effectively monitor soil salinization changes for the study area.