Abstract: It is important to analyze the reliability level of multi-components in a structural system accurately and efficiently. Monte Carlo method and response surface method are usually used in such reliability analysis. However, the number of structural analysis in Monte Carlo method depends on the value of the reliability index, which usually requires large computation cost. The number of structural analysis in response surface method depends on the number of components, which also needs significant computation cost when the number of components is large. A reliability method for multi-components in a structural system based on the adaptive point estimate method and the principle of maximum entropy is proposed. In this method, the upper limit of the required number of structural analysis is irrelevant to the number of components, and the computation process is easy to implement without iterations. Firstly, the first four moments of each component are calculated based on the combination of adaptive delineation of cross terms and bivariate dimensional decomposition. Then, the principle of maximum entropy is induced to evaluate the reliability index of each component according to the first four moments. Finally, several cases are investigated to compare the accuracy and efficiency of the Monte Carlo method, the response surface method and the proposed method. The results demonstrate that the proposed method has significant advantages in efficiency when compared with Monte Carlo method and response surface method, and can be implemented with satisfactory accuracy for engineering problems.