Abstract. Scale, context, and heterogeneity have been central issues in geography. From a quantitative standpoint, accurately identifying the scale and context at which geographical processes operate and capturing their spatial heterogeneity have been challenging tasks. Despite various prominent developments in spatial modeling literature, there is a lack of models for separating individual- and group-level spatial processes that may also exhibit spatial heterogeneity. Understanding this difference can better inform us about how individuals are separated from or influenced by higher-level contexts. In this regard, we propose Multilevel Geographical Process Models (MGPMs) to simultaneously incorporate both individual- and multi-level spatial process heterogeneity. We demonstrate the performance of the model using Monte Carlo simulations and Compare it against Multiscale Geographically Weighted Regression and Multilevel models.