The spatial distribution of forest canopy cover is a critical indicator for evaluating the forest productivity and decomposition rates. With the Wangqing Forest Region in Jilin Province of China as the study area, this study first estimated the forest canopy cover using spaceborne LiDAR ICESat-GLAS waveforms and Landsat-TM multispectral images, respectively, and then GLAS data and TM images were combined to further estimate forest canopy cover by using multiple linear regression and BP neural network. The results showed that when the forest canopy cover was estimated with single data source, the determination coefficient of model was 0.762 for GLAS data and 0.598 for TM data. When the forest canopy cover was estimated by combining GLAS data and TM data, the determination coefficient of model was 0.841 for multiple linear regression, and the simulation precision was 0.851 for BP neural network. The study indicated that the combination of ICESat-GLAS data and Landsat-TM images could exploit the advantages of multisource remote sensing data and improve the estimating accuracy of forest canopy cover, and it was expected to provide a promising way for spatially continuous mapping of forest canopy cover in future.