Abstract: The global reliability evaluation of structures has always been a difficult and hot spot in the field of reliability analysis. It first classifies the global reliability problem of structures and provides a unified description of the corresponding performance function. Combining with the proposed effective dimension two-step analysis method with the conjugate unscented transformation method, an improved statistical moment point estimation method is developed. Integrating the principle of maximum entropy and the improved statistical moments point estimation method, a new reliability analysis method is proposed, which is suitable for two types of global reliability problems. The proposed method is verified through two numerical examples. The results show that, at the same level of accuracy, the computational efficiency of the proposed method is 2.3 to 2.6 times higher than that of the conventional trivariate dimensional reduction approximate statistical moment point estimation method, and that the proposed method has a high level of accuracy with the maximum relative error less than 2%, which is suitable for the global reliability assessment of structures.