The pavement maintenance is a critical component of the pavement life cycle. The effective pavement maintenance needs the pavement performance under do-nothing scenario as well as under various treatment choices. The performance of a pavement is characterized in terms of the various distresses viz., functional distress and structural distress. When a pavement is intervened for maintenance, each treatment offer varied levels of effectiveness for different distresses. These are expressed in deterioration models and treatment effectiveness models. As pavement behaves as a system, the system dynamics concept can be adopted to study the interaction of deterioration progression and treatment effectiveness on traffic loading over a period of time. Integrating the above components with the agency cost, a system dynamic model is formulated and implemented using system dynamic software POWERSIM. Optimization is performed using evolutionary principle in the software. This study identifies the minimum agency cost required for optimal scheduling of pavement maintenance for a planning period of 9 years. © 2018 Taylor & Francis Group, London.