Abstract: Urbanization, growth of urban areas, is a process that has been growing rapidly during the last two decades. This phenomenon affects aerobiological, economical, industrial, ecological processes, social control, and the family. Hence, the prediction of the urban area extent has an important role in the future decision of the municipality. Multi-Agent System (MAS) is a proper tool for simulation and modeling process, which has been used for solving different types of spatial and non-spatial problems. In this article, we use MAS for urban simulation in the rural area around Tehran, which is the most populated and the fastest-growing city of Iran. In this paper, the behavior of three groups of agents: environmentalist, industrialist, and resident are simulated. These three groups are the dominant and influential population in the formation of urban texture. In this research, the behavior of these three groups of agents is specified, according to a series of map layers, such as slope, aspect, soil type, distance of urban areas, roads, and so on. The Belief-Desire-Intention (BDI) architecture of agents is used for the simulation, which is defined based on some variables, functions, and coefficients. The simulation is carried out based on two different interaction scenarios: Rational and Nash-Equilibria. The future urban area is predicted by a combination of MAS and spatial urban area. To evaluate the proposed model, the comparison of the predicted area is made at different times and scenarios. The results of implementation in different scenarios show that the residents of the study area follow the Nash-Equilibria interaction with Kappa Coefficient accuracy of 0.8104.