In our recent book 'Spatial Simulation: Exploring Pattern and Process' (co-authored with George Perry) we propose a classification of dynamic spatial models into broad process categories of aggregation-segregation, movement, and growth. I revisit these categories and show how thinking about process models in this way leads to ‘building block’ spatial models. Such building block models offer a number of advantages in helping us to build and to understand complicated simulation models. Further, these building blocks may provide a way forward in the development of more dynamic, process-based, geographical information systems.