On scientific realism, mature scientific theories are (approximately) true, and in virtue of this (approximate) truth they are successful. According to the substantialist accounts of scientific representation, scientific models enable cognitive agents to reach epistemic virtues because they represent the world. This article aims to scrutinize the question of whether scientific realism is consistent with the three well-known approaches to scientific representation, namely similarity-based, agent-based, and hybrid views. To do so, it will be argued that these approaches are, at least prima facie, orthogonal to scientific realism. After addressing the ways in which they become related, it will be argued that scientific realism is inconsistent with instances of all three approaches, provided certain assumptions are taken into account. -- -- -- -- -- -- -- -- --- -- --- --- --- --- --- --- -- -- --- -- --- -- --- --- -- -- --- --- -- -- --- -- --
)