Soil salinization is a most important factor to influence agriculture in arid area. There is a dire need to monitoring the soil salinization information based on Chinese own satellite data. In the study, a remote sensing classification method of soil salinization land based on knowledge discovery and decision-making process that is benefit to arise classification precision through a decision tree method has been constructed. The result shows that near 80% area of the entire county has suffered soil salinization with different grade. The percent of heavy, middle and light grades soil salinization area are 11.29, 20.53 and 49.01 respectively.