Abstract: Due to the complex and varied shapes of curved stainless steel buildings， current positioning measurement methods lack error compensation， resulting in lower positioning accuracy and larger errors for curved stainless steel buildings. Therefore， a method for measuring the positioning accuracy of curved stainless steel buildings based on laser trackers is proposed. According to the measurement requirements of curved stainless steel buildings， a laser tracker with high precision， wide-range measurement capability， and good tracking performance is selected. The specific parameters include measurement range， measurement accuracy， adaptability to working environment， etc.， and the complex curved stainless steel building that needs to be measured is taken as the positioning measurement target. The building itself is used as the origin of the coordinate system. Parameters such as the distance from the reflective target to the measurement point， and the horizontal and vertical angles output by the tracking head are set to establish the building coordinate system. A tracker installation system is constructed， comprising components such as a laser tracker， dual-frequency laser interference optical path， lens groups， integrated detection optical path， tracker base stations， signal control boards， and real-time displays. Raw measurement data is obtained in the form of a measurement network. The distance between the measurement point and the observation point was measured using a laser interferometer （IFM） and a laser absolute rangefinder （ADM）. Based on the principle of multilateral measurement， a mathematical model was established to control the laser’s emission and return distance， which facilitated the acquisition of raw measurement data. The collected laser images were digitally processed and stored in a database. The fixed vibration displacement and amplitude of the gimbal were calculated and fused with the measured values to eliminate gimbal errors. Filters were added to the laser tracker to improve resolution， and both software and hardware filtering operations were performed to avoid signal errors. At the same time， a virtual offset spot was set for spot compensation. After error analysis and compensation， the processed data were organized into positioning data outputs. Based on the output data， the positioning accuracy of curved stainless steel buildings was evaluated to ensure the measurement results met the design requirements. The test results showed that this method met the current requirements for tracking distance trajectory and tracking deviation curve， and exhibited a high overall positioning accuracy， achieving high-precision positioning measurement of curved stainless steel buildings.