This study developed and discussed a framework for characterising rationales for using statistical modelling from a mathematical modelling perspective based on a systematic literature review. We used this framework to provide an overview of the distributions of these rationales in the analysed studies, focusing on statistical modelling. The results identified three categories of rationales for using statistical modelling, namely, competency, content, and socially oriented types. This work discussed how the conceptualised rationales for using statistical modelling in the teaching and learning of statistics can guide the initial step in task design and curriculum development related to statistical modelling. These rationales may also serve as a common point of departure for discussions and collaboration between the mathematics and statistics education communities on the use of models and modelling.