Spatio-temporal modeling for urban applications has received special attention lately. Due to the recent advances in computer and geospatial technologies, the temporal aspect of urban applications which was ignored in conventional systems, is under consideration nowadays. This new interest in spatio-temporal modeling, in spite of all its deficiencies, has brought about great advances in spatio-temporal modeling and will enhance the urban systems dramatically. This paper investigates two different viewpoints in spatio-temporal modeling for urban applications. The first category involves CA based modeling, agent based modeling, Artificial Neural Networks modeling and fractal based modeling. These models which have been widely used in simulating complex urban systems are here distinguished as complexity models. The applications of these approaches in modeling complex urban systems are comprehensively reviewed and advantages and weak points of each model are depicted. On the other hand, temporal GIS models as another approach for spatio-temporal modeling are briefly reviewed. Eventually the conceptual differences between these two categories are mentioned to aid modelers to mindfully select the appropriate models for urban applications.

Keywords