The reliability of mining systems is generally low due to their harsh working conditions. Currently, efforts for improving mining system reliability are often made in isolation. This practice could substantially limit the effectiveness of the efforts on overall reliability improvement of the mining system. To enhance the overall reliability of mining systems, an integrated improvement approach is necessary. In this paper, we developed a framework for integrated mining system reliability improvement to address this issue. In this framework, there are five major components including data integration, business process integration, hardware integration, software integration, and analysis/decision integration, but we only focus on the integrated reliability analysis which is important to the analysis/decision integration. The reliability analysis considers the interactions between machines, and the impacts of design, operation, maintenance, automation and working environment on the overall system reliability. These multiple interactions present a big challenge to accurate reliability prediction. In this paper, we for the first time systematically investigated integrated reliability analysis approaches for dealing with this challenge using novel models and methods, including covariate hazard models, intelligent reliability prediction approach, and complex system modeling methods. While these models and methods have found some successful applications in other industries, they in general have not been effectively used for the reliability analysis of mining systems. Our study results show that the system integration approach is applicable to mining systems and can be used for developing a computer aided integration system for the implementation of the integrated reliability improvement approach.