Abstract: Light detection and ranging（LiDAR）, as an active remote sensing technology, is able to obtain information on the spatial structure of forests by emitting laser energy and receiving the return information, however, when used alone there is a scanning blind spot and a complete 3D point cloud of forest trees cannot be obtained. Based on this, this study proposes a method for estimating the structural parameters of a individual tree by fusing UAV and TLS LiDAR point clouds, and realizes point cloud fusion by using a registration method based on ground features and tree position relationships. Based on this, an improved K-means hierarchical clustering segmentation algorithm is proposed to complete the individual tree segmentation, and then based on the segmented individual tree point cloud, the axis-aligned bounding box algorithm and the least squares fitting circle method are used to extract the height of the individual tree and diameter at breast height respectively, and finally estimate the biomass of a individual tree through the biomass allometric growth equation. The research results show that the Coefficient of Deter mination（R~2） of tree height, diameter at breast height and individual tree biomass of Quercus mongolica sample plots are 0.84, 0.93, 0.91 respectively, and the root mean square error（RMSE） of individual tree structure parameters are 0.75 m, 0.96 cm, 26.31 kg/plant respectively. The R~2 of tree height, diameter at breast height and individual tree biomass in the Pinus sylvestris plot are 0.92, 0.96, and 0.95, respectively, and the corresponding root mean square errors are 0.43 m, 1.06 cm, and 26.12 kg/plant, respectively. The fusion of UAV and TLS LiDAR point cloud provides a reliable data basis for the rapid and complete acquisition of forest structure information, and provides a strong technical support for the deep forestry application of joint multi-source LiDAR technology.