||Road assets are constructed to accommodate the needs of people and to serve various functions associated with such needs over their service life. In this study, the principles and framework of a performance-based road asset management system (PBRAMS) are established to practically and efficiently maintain desired services over the service life of road assets under limited resources. The main components of the PBRAMS, such as road network hierarchy, asset performance measurement, maintenance cost and effect modeling, asset performance modeling and asset management optimization, are discussed. One major function of the PBRAMS is the road asset management optimization model in terms of appropriate asset performance modeling and predetermined management policy in order to make optimal asset management plans. In this study, asset management optimization models have been developed with illustrative examples on the basis of different asset performance models including time-based renewal process, discrete-time Markov process and discrete-time semi-Markov process. The results are obtained and compared to show the applicability of different asset management optimization models. Moreover, this study considers the social impacts in the road asset management. Major social cost components are identified and general cost models are built. A multiobjective optimization model has been developed to incorporate social costs into the road asset management optimization and an example is given to illustrate the use of this multiobjective model. Systematically incorporating key road asset management components, the PBRAMS provides a solid foundation for a road agency to predict asset performance over the service life of its managed road assets and to optimize the short- and/or long-term asset management plans.