The notion of incommensurability has provided a rationality criterion for the development of scientific theories, as well as some insight into theories developed by students while learning science. However, the relationship between the multiple models held by students and incommensurability requires further discussion. We present the results of empirical work that investigated the multiple models of the structure of the matter held by university students and we analyze these results using the notion of incommensurability. We also point out implications in the construction of students' scientific models as they move forward in their careers.