The lack of systematic geological work is an essential reason why underground coal gasification (UCG) has not been industrialized for a long time. Building a scientific index system and favorable area evaluation technology for the UCG site selection is the key to breaking through the geological bottleneck. Aiming at the problems of the single index weight determination method, intense subjectivity, and poor reliability of current evaluation models, we put forward an evaluation modeling methodology for the UCG site selection using the combination weighting method with the game theory. The factors of coal resource conditions associated with the potential risk of UCG are systematically analyzed. From the six dimensions of the geological structure, hydrogeology, seam occurrence, coal properties, reserves, and roof lithology, 23 key factors were selected as evaluation indexes to construct a hierarchical model composed of the target layer, category index layer, and index layer. The influence of each index on UCG and its reasonable value range were systematically analyzed. The evaluation index system for UCG site selection was formed. The improved analytic hierarchy process (AHP) was adopted to sequence indices and determine their subjective weight. And the variability, conflict, and information amount of the index data were analyzed by the CRiteria Importance Through Intercriteria Correlation (CRITIC) method to calculate the objective weight. Then, the subjective and objective weights were combined through game theory. On this basis, fuzzy theory was employed to calculate the membership of indices and construct the fuzzy comprehensive judgment matrix. The evaluation model of the UCG site selection was applied to the suitability evaluation of resource conditions of UCG pilot projects at Zhongliangshan (ZLS), Huating (HT), and Shanjiaoshu (SJS) mines in China. The result shows that the resource conditions of HT are the best, followed by ZLS and, finally, SJS, which are consistent with the actual running effects of the three UCG pilot projects. It indicates that the evaluation model can provide a scientific theoretical basis and reliable technical support for the UCG site selection.