Comprehensive evaluation is widely used in many fields, and the evaluation results are an important basis for the next decision. In order to obtain reasonable and reliable evaluation results, the evaluation indicator system must be able to scientifically, comprehensively and accurately reflect the object and purpose of the evaluation, and all indicator data must be authoritative, accurate and properly processed. Based on the above requirements, the construction of resilience evaluation indicator system and data processing method are put forward, the specific use steps are combed out, and the calculation process and results are given by taking China’s provincial resilience evaluation as a case study. The construction of the evaluation indicator system includes the primary selection of indicators based on evaluation objectives and existing researches, the optimization selection of indicators based on data acquisition, and the screening selection of indicators based on KMO test statistics and Measure of Sampling Adequacy (MSA). The data processing is carried out after the optimization selection of indicators, including the time sequence three-dimensional data table, the consistency processing of the indicator type, the non-negative processing and the dimensionless processing. Based on the processed time sequence data table, the eigenvectors corresponding to the positive maximum eigenvalues can be calculated, and the elements in the eigenvectors are the weights of each indicator.