Abstract: In response to the problem of less method for accuracy evaluation of point cloud data, this study proposes a comprehensive accuracy assessment method integrating points, lines, and surface-based analyses for point cloud data in complex terrains.By analyzing the three types of errors in point cloud data accuracy assessment, a point cloud data accuracy assessment method and a point cloud data quality evaluation system are established.Then, by analyzing the error propagation law, the theoretical accuracy formula of point cloud data is derived.Subsequently, the data results of points, lines, and surfaces extracted from point cloud data are compared with the data collected by traditional methods for single-point and overall accuracy analysis, and a comprehensive accuracy evaluation system based on points, lines, and surfaces is established, which combines single-point accuracy and overall accuracy, internal consistency accuracy and external consistency accuracy.Experimental data analysis using this method shows that the accuracy of point cloud data is up to 0.07 m under influence of terrain and up to 0.02 m under influence of landform based on point accuracy analysis; the difference in cross-sectional area is generally within 0.6% based on line accuracy analysis; and the volume difference of different meshing methods is within 1% based on surface accuracy analysis.The experimental results demonstrate that this method has reference value for the accuracy assessment of point cloud data under complex terrain and can be applied in actual project engineering.