Abstract: In this paper, the fifth desertification monitoring area in Zoige county was selected as the research area, and the NDVI was extracted from Landsat 8 images in early October of 2013, 2015 and 2017. The vegetation coverage (FC) was inversed by the dimidiate pixel model. After classifying the NDVI and FC, the area transfer matrix of each level was analyzed, and the correlation analysis between NDVI , FC and the desertification degree was also analyzed. The results showed that: (1) The NDVI and FC showed a significant increasing trend from 2013, 2015 and 2017, with the median value of NDVI increasing from 0.4775 in 2013 to 0.5374 in 2015 and reaching 0.5921 in 2017; the median value of FC increased from 0.5305 in 2013 to 0.5971 in 2015 and reached 0.6578 in 2017. (2) In the area transfer matrix of NDVI and FC levels, the specific transformation direction was the same, mainly to the direction of higher vegetation, which was shown as: the medium level to the higher level; the lower grade to the medium grade; the low level to the lower level. (3) The NDVI and FC levels were significantly positively correlated with the desertification degree monitored in 2015, indicating that the classification of NDVI and vegetation coverage reclassification grades was basically in line with the current situation of desertification vegetation coverage in the study area. The research methods and results could provide data and technical support for desertification monitoring.