Even though education is a widely acknowledged modelling purpose in the agent-based social simulation literature, we currently lack a sufficiently detailed and robust understanding of how computational models can be used for the training of new entrants and the upskilling or reskilling of current practitioners. Considering the potential benefits of readily actionable guidelines and recommendations, this text briefly addresses four macro-level dimensions that warrant consideration by those interested in incorporating agent-based simulation into their teaching practices: (i) the expected learning outcomes, (ii) the pedagogical approach, (iii) the curriculum organisation, and (iv) the institutional setting. The goal is not to provide an overarching rationale for simulation-based learning or detailed implementation procedures, but, rather, to suggest a starting point for practitioners (especially those unfamiliar with curricular and pedagogical issues) that are wondering how to transfer their professional expertise into the classroom. Best teaching and learning practices, the discussion will show, must transversely acknowledge and accommodate teaching and learning needs and restrictions across these four dimensions.