This study utilizes system dynamic approach to discuss activities of land development planning. To operate, a computer simulation model is build at tempt to describe some possible developments of a land-use plan and the planning behavior. The simulation model assumes that a planner is going to deal with a gap between demand and supply of certain kind of land, and trying to solve the problem by coverting other kinds of land for the gap. However, this action further influences the planner's original expectation. Simulated outcome indictates that the major variables and causal relations including in the model are able to produce enough dymanics which contribute to model behavior. Also, simulated outcome does not violate logical and theoretical assumptions made in the model. After validity and reliability are secured, a number of scenario analyses are design to test policy applicability of the model. First, to see if a planner changes her/his value of different kinds of land what would happen to supply and demand of land use. Second, to see if a planner changes her/his criteria of setting the goal of landconvert. The results show some complex phenomenon which should be concerned in land development project.