For the sensitive of distribution parameters of random variables, system reliability research should be focus on identifying the significant failure modes, while failure probability as a quantitative reference index, only combining both of them can better estimate system reliability. On the basis of load incremental step theory, a method for identifying the significant failure modes which could more accurately simulate the real failure process for considering the change of bearing capacity and spatial position is proposed in this paper. Then, the corresponding program is prepared with APDL on ANSYS platform. On the one hand, the efficiency and accuracy of identification is improved, on the other hand, the usefulness of this method is enhanced. Finally, the reliability of a space grid engineering project is analyzed and discussed by this method; moreover, the system reliability index β is calculated as well. These analysis results are helpful to provide scientific basis for structural optimization, structural testing, and structural reinforcement and so on.