Abstract This article investigates an overlooked question: what kind of methodological sensitivity is necessary when using Laurent Thévenot's sociology of regimes of engagement (SRE) in ethnographic research? Thévenot's call to study proximate forms of engagement ‘below’ the public correlates with the recent turn to non-representational methodologies in ethnography. Thus, we discuss the methodological consequences of the ontological commitment of SRE through a dialogue with the epistemological arguments of non-representation, which we take to include sensory, affective, and embodied ethnographies. We do so by zooming in on three key aspects of SRE: plurality of different regimes of engagement, investment of people in their environment to secure goods, and versatility between trusting and doubtful engagements. We argue that while SRE and non-representational ethnography share a concern for unspoken, affective, and embodied practices, SRE requires a different sensitivity towards the dynamics between the representational and non-representational.