Abstract: The 3D laser scanning technology has provided a new solution for the monitoring of settlements and de? formations in metro tunnels under complex conditions by acquiring the 3D coordinates of target objects in a fast and highly precision way. This paper combines the factors affecting the precision of scanner points at the construction site and designs a quasi-level experiment of scanning errors using the orthogonal experimental method, which takes no or any vehicle running load conditions, different horizontal ranges and different measurement environment conditions as the factor levels. By analyzing the variance, contribution rate and the indicator-factor level, the influence degree of each factor on scanning errors is obtained. Based on the analysis of the experiment results, it establishes two error correction models by using the baseline comparison model and in conjunction with the actual monitoring requirements of project engineering. The results show that horizontal ranging and vehicle running load are significant factors affecting scanner errors. Combined with the established error correction model, the systemic correlation between the corrected point errors and the scanner ranging is significantly reduced, and the point coordinates meet the precision requirements for later point cloud splicing and monitoring points in the specifications, which provides a theoretical basis for the application of 3D laser scanning technology in the monitoring of settlements and deformations of metro tunnels.