Abstract: Naiman county is one of the counties with serious land degradation. In order to figure out its degradation type, distribution and degree and monitor its degradation process and trend, the land use/coverage change, sandy area ratio in a pixel, composite vegetation index, and soil moisture were selected as four indexes of land degradation evaluation system. All the four indexes were retrieved from TM images. Then the judge matrix of four indexes were constructed using analytical hierarchy process, the eigenvector of judge matrix was calculated to obtain the weight of four indicators. Finally the model of land degradation estimation system was built with four weights. According to the model, the land degradation of Naiman county was classified into 5 levels, no degradation, slight degradation, medium degradation, bad degradation and serious degradation. The degradation evaluation results demonstrated that land degradation of Naiman county had a improvement in some local area, but overall worsening during 1987-1999, while the land degradation has been constrained and degradation treatment has gained some effect since 2000. The research is of some importance and guidance to the degradation treatment and ecological construction in Naiman county.