Abstract: Generally, dense urban areas suffer from local parking shortages: a limited supply coupled with a diffuse demand in both space and time. According to this plan, additional vehicles will be searching for available parking spaces, particularly during peak hours. Searching for parking disrupts the economic, environmental, and social sustainability of the city. The purpose of the paper is to optimize parking management, by guiding drivers who request a vacant space. Multi-agent systems can be used as an appropriate modeling approach for the development of such a smart parking guidance system. In order to verify and validate the model, simulations were performed in Tunis city-center. The numerical results show that smart parking guidance systems can manage parking areas more efficiently. It can reduce energy consumption and carbon dioxide emissions, increase revenues for parking managers and satisfy drivers requirements.