Geographic research necessitates different types of models that aim at reducing the complexity of reality, enabling us to analyze its representations. Models allow us to infer exploratory and confirmatory findings about causal links between different geographic conditions, different social, political, and economic actors, their behaviors, and land use and its dynamics. However, the spatial, temporal, and attributive nature of data complicates the appropriate model choice for specific research purposes. This conceptual paper keeps track of model construction in the context of land use and land use change. It reveals different issues where particular caution is required, such as spatial and temporal dependency and heterogeneity, scale dependency, and bidirectional causality. With geographic research being increasingly under the influence of econometric methods, we contend that it is necessary to provide intermediary qualitative layers of evidence between conceptual and, especially, empirical and simulative models of land use and land use change to ensure robust findings about causal relationships between observed variables.