Snow cover area is one of the important input parameters in hydrology and meteorology models. Getting real-time information of large-scale snow cover through remote sensing can make up the lack of scope in space by ground-based observations. As medium-resolution imager (MODIS) data has multi-spectral as well as high time resolution, more and more applications were introduced to the snow. Based on summation of the existing methods, the paper analyzed the advantage and disadvamtage of the present methods and established a new snow cover informatioon, linear relation model by NDVI (normalized difference vegetation index) and NDSI (normalized difference snow index). The accuracy of the model has been verified. The results showed that in areas with high vegetation coverage, the method can reduce the misclassification of 8.7% of the pixel.