Abstract The combination of the three-dimensional (3D) target mirror and laser interferometers can achieve 3D precision measurement, and the orthogonality between the target mirror has a significant impact on the accuracy of the measurement system. In order to ensure the accuracy of the measurement system, it is necessary to measure the orthogonal deviation angle of the 3D target mirror and evaluate the uncertainty to verify whether the machining accuracy of the target mirror meets the requirements. In this paper, to test the orthogonality of the X-Y reflection planes of the 3D target mirror, which is glued by three plane mirrors, the evaluation methods of measurement uncertainty of the orthogonal deviation angle θxy are studied. The measurement model of uncertainty evaluation is built based on the quantitative characteristics analysis method, and the evaluation results of measurement uncertainty are obtained based on the adaptive Monte Carlo method (AMCM). Moreover, the evaluation results are compared with those evaluation results obtained respectively based on the traditional assessment method of measurement uncertainty (called as GUM method) and Monte Carlo method (MCM). Through the comparison of the evaluation results, it can be concluded that the best estimate of the orthogonal deviation angle θxy is about -3.09”, the corresponding standard uncertainty is about 0.69”, and the machining accuracy of the X-Y reflection planes of the target mirror meets the requirements. Moreover, the AMCM evaluation results is reliable, and AMCM is more convenient and efficient than GUM and MCM methods. The research work provides valuable reference for solving practical measurement evaluation problems.