Traditional reliability analysis is based on probability theory where the distribution of parameters is precisely known. However, the distribution of several uncertain variables cannot be accurately obtained in real practice due to the insufficient information. Occasionally, random variables and interval variables may obtain simultaneously, and it is inappropriate to continue to adopt the probability-based reliability analysis methods. The existence of interval variables also makes the reliability analysis more difficult. To overcome these shortcomings, an effective hybrid reliability analysis method is proposed for structures in this paper. The mixed uncertainty model is divided into probabilistic analysis loop and interval analysis loop based on the decoupling method. Then the conjugate finite step length algorithm is developed for probabilistic analysis, which significantly improves the computational cost. Two numerical examples are provided to illustrate the accuracy and efficiency of the proposed method.