Fractional Vegetation Cover (FVC) describes the distribution and growing condition of vegetation on land surface ecosystem, and it could be estimated in regional and global scale with remote sensing techniques conveniently. The multispectral sensor equipped on China's GaoFen-4 (GF-4) geostationary satellite which was launched in December 2015 has a high spatial resolution of 50 m, thus makes it a powerful data source for vegetation monitoring and FVC estimation in large spatial scale timely. In this paper, we conducted a FVC estimation experiment in Northeast Inner Mongolian, China with GF-4 data acquired in August 2016 based on 2 most widely used method, and the estimation results were validated by a simultaneous Unmanned Aerial Vehicle (UAV) measurement afterwards. Our results showed that the vegetation index method with pure endmember pixel of soil and vegetation NDVI (Normalized Difference Vegetation Index) value extracted in a 2-dimensional spectral space was most accurate with the near ground UAV measurement results, and it would be a promising method in GF-4 FVC estimation in northern temperate semi-arid area of China in the future.