The goal of the research presented was to create the tool that could help to define, predict and assess ex-antem sustainability of plans for development of an urban territory. The model is intended for simulation of urban development dynamics on local scale (urban municipality). The tool uses model representation of urban environment that reflects the externalities that stem out from various types of land uses and relative accessibility of each site in the represented area. The dynamics of the model is above all influenced by positive and negative externalities from uses in immediate neighborhood, spatial accessibility of the site, spatial accessibility of the infrastructure, spatial and functional constraints of spatial development declared as by-law regulatives, and transformation cost connected to change from one function to another. Cellular automata was selected as the best way to represent the spatial influences of various land-uses. Space is represented by one-dimensional or multi-dimensional grid of cells. Each cell has a certain number of neighborhoods depending on the defined size of neighborhood. In the model, each cell represents one type of land-use. The change from one land-use type to another is determined by the transition rule. Multiple factors (neighborhood effects, relative accessibility and cost of land-use type change) enter the transition rule leading to the best possible use from the point of view of a single cell. The case of a suburban city Říčany next to Prague is used for the model testing and calibrating. The issues of model variable sensitivity, output validity and the model applicability are discussed.