Abstract: LIDAR is a new technology for obtaining instantly the accurate 3D information of ground and objects.A slope-based planar-fitting filtering algorithm of LIDAR point cloud is presented based on the analysis of the spatial distribution feature of LIDAR point cloud.This algorithm is used for the experimental research of feature extraction of airborne LIDAR point cloud.The experimental result shows that this algorithm is able to fit effectively continual and smooth horizontal planes,slope planes and vertical planes,and has preferable effects in the feature extraction of DTM,building tops and vertical walls.