This paper attempts to explore some of the contributions that philosophy of science can offer to the didactics of the empirical sciences. Initially, a brief characterization of the so-called “school science” based on theoretical models is provided. Then, we review the meta-theoretical developments around the Kuhnian notion of “empirical comparability” with the aim of identifying possible contributions to the meta-theoretical foundation of this school science and, more generally, to science teaching based on modeling. At the end of the paper, an activity for the teaching of Daltonian stoichiometry built on these reflections is proposed as an example.