In this presentation I outline my research proposal “Explanatory Reasoning: a Mechanistic Perspective” for the research project “Methodological and Epistemological Analysis of Scientific Reasoning Processes”. The overall aim of my proposal is to develop an account of explanatory reasoning in which epistemic virtues that are invoked in explanation and theory choice are clarified conceptually, and their explanatory import is explicated. I aim to develop this account by expanding on recent work on norms of mechanistic explanation. I introduce the topic, aims, background, and approach of my proposal. I end by identifying research challenges. I illustrate the approach by describing a first line of research that I aim to undertake to elaborate norms of mechanistic explanation further. Central to this research line is disambiguating the concept of role function in mechanistic explanation. The concept of role function is a key concept in the construction of mechanistic explanations. The explanatory import of role function ascriptions is explaining how entities and their behaviors contribute to overall behaviors of mechanisms. The relevance of role functions in evaluating the adequacy of mechanistic explanations is, however, less emphasized. Rather, a variety of norms by which mechanistic explanations are to be evaluated have recently been proposed. For instance, the norm of ‘mutual manipulability’ to assess the relevance of a putative component part in a mechanism: a part/entity is a component in a mechanism if one can change the behavior of the mechanism as a whole by intervening to change the behavior of the part/entity, and vice versa. Now, if a part is relevant for the production of a behavior of a mechanism as a whole, it can be ascribed a role function, since it contributes to the production of this overall behavior. However, I have argued elsewhere that the mechanistic concept of role function is ambiguous, and can be disambiguated in two different senses: on the one hand, behaviors of component parts that make some crucial contribution to the overall behavior of a mechanism and, on the other, the role fulfilled or contribution made by behaviors of component parts. For instance, the heart’s behavior of ‘pumping blood’ can be ascribed a function in particular contexts, as well as the role fulfilled/contribution made by the heart’s pumping of blood; for instance, ‘delivering oxygen and nutrients to the rest of the body’. Yet, the ‘mutual manipulability’ norm does not distinguish between these two senses in which a component part can be relevant to the behavior of a mechanism as a whole. I argue that by connecting norms of mechanistic explanation to these disambiguated senses of mechanistic role function, precision can be gained on what is precisely meant by such norms. I expect that connecting norms of mechanistic explanation to role functions also contributes to clarifying other relevant issues, such as the (disputed) distinction between constitutive and causal relevance in mechanisms.