The landscape or regional scale of analysis poses certain challenges and new possibilities for better understanding of space and time-related processes and changes. These changes can be caused by exogenous processes (e.g. climate change, dynamics on the global market, or technological advances), or may be a result of endogenous regional processes (e.g. dynamics of land use changes). The regional scale is regarded as an area from about a 100 km2 up to 1000 km2. The main problem is that at this scale we can still distinguish the signals from a variety of local processes and cannot ignore them by averaging, while the sheer number of these processes replicated over a large enough territory makes the system extremely complex, structurally diverse, and ecologically heterogeneous. This complexity is further exacerbated by enhanced uncertainty in data that spans time and space. To cope with these problems we suggest an approach that promotes models of intermediate complexity. An important feature of Regional Models of Intermediate Complexity (REMICs) is that they are characterised by a lower degree of detail in the description of process dynamics, but a higher number of interacting components. The implementation of REMICs calls for specific modelling tools that can handle the spatial heterogeneity of GIS and can offer analytical capabilities of both process-based and statistical modelling. One such tool, the Spatial Analysis and Modeling Toolbox (SAMT) is introduced. SAMT is an open source package that can be used to develop decision support systems (DSS) that can help planners and decision makers better understand the complex reactions of the landscape to various forcings.