With the rapid development of highway construction and formation of the highway network in China, the man-agement of pavement maintenance and rehabilitation (MR) activities has become important. In this paper, four discrete optimization models are proposed for different parties involved in the management system: government, highway agent, con-tractor and the common users. These four optimal decision models are formulated as linear integer programming problems with binary decision variables. The objective function and constraints are based on the pavement performance and prediction model using the pavement condition index (PCI). Numerical experiments are carried out with the data from a highway system in Sichuan Province, which show the feasibility and effectiveness of the proposed models.