Abstract: A data-driven yaw steady-state error calibration method was proposed, aiming to improve power generation efficiency without extensive hardware modifications. Taking a 2 MW horizontal-axis wind turbine as the research object, a steady-state error calculation method based on data acquisition and monitoring control system (SCADA) was established, providing a quantitative reference for error calibration. A nacelle-mounted lidar was installed on the target turbine to participate in calibration verification. By analyzing and comparing multi-source wind direction data, a strategy for lidar participation in steady-state error calibration was developed. Considering the error calibration references calculated from both SCADA and lidar data sources, the method was applied and evaluated in the field on a commercial turbine. Results show that the yaw control accuracy and power generation efficiency of the turbine have been improved significantly after calibration. The proposed method can be used for technical upgrades to enhance the performance of wind turbines.