Desertification is a worldwide concern and the assessment of aeolian desertification has become one hotspot in global ecosystem research. In this paper, hyperspectral data acquired from modular OMIS-I imaging spectrometer, combined with ETM data and field survey data, was used to assess the aeolian desertification in Korqin Sand, Inner Mongolia, China by pixel-level. The results indicated that hyperspectral image, combined with ETM image and little field works, is capable to monitor and assess desertification through quantitative retrieval of assessing parameters directly from hyperspectral data or indirectly from the encoding map by visual interpretation of hyperspectral image and ETM image. For the retrieval of vegetation biomass and coverage, polynomial fit curve is suitable to regions where shrubs and grasses coexist, while linear fit curve is suitable to single vegetation type and was highly restricted by region. The retrieval of surface soil water content based on soil thermal inertia is suitable in flat terrain and sparse vegetation, and it can resist vegetation disturbance. The algorithms for numerical evaluation and quantitative retrieval for hyperspectral image are also practicable for aeolian desertification in Korqin Sand, China.