This paper discusses a lesser-known criticism of the phenomenological method and its use of imaginative procedures, which can be found in ethnomethodology, a minor direction of contemporary empirical sociology. Often seen as an empirical extension of Alfred Schütz' phenomenological sociology, ethnomethodology shares with phenomenology the ambition of clarifying structures of our social life-world on ground of our intercorporeal mutual undertstanding in concrete interaction. However, it mainly differs from phenomenology – and this relates to the objection I have in view– in stressing that the decisive details of these processes of interaction that ground mutual understanding (gestures, bodily expressions, or specific stances in casual conversation) are „not imaginable, they can only be found out” (Garfinkel 1996, Sacks 1992). My purpose here is to expand on what such a claim of unimaginability entails for phenomenology. I will do so by first delineating this line of criticism from other objections raised against the imaginative procedure within or without the phenomenological camp (Lohmar 2005, DeSantis 2011, Ferencz-Flatz 2011, Adorno 2013 etc.) and by showing how the question thus posed departs from the core philosophical debates concerning imaginabilitiy in the Kantian tradition. Further on, I offer an in-depth interpretation of the aforementioned objection in order to outline not only its methodological implications for the discussions concerning the need for empirical procedures in phenomenology, but also in order to show that it involves key insights for understanding interaction, which phenomenology can valorize.