In this essay I intend to flesh out and discuss what I consider to be the groundbreaking contribution by the German historian and theorist of history Reinhart Koselleck to postwar historiography: his theory of historical times. I begin by discussing the view, so prominent in the Anglophone context, that Koselleck's idea of the plurality of historical times can be grasped only in terms of a plurality of historical periods in chronological succession, and hence, that Koselleck's theory of historical times is in reality a theory of periodization. Against this interpretation, to be found in works by Kathleen Davis, Peter Osborne, and Lynn Hunt, among others, I will argue that not only is Koselleck's theory of historical times, or, with a more phenomenlogical turn of phrase, his theory of multiple temporalities, not a theory of periodization, it is, furthermore, a theory developed to defy periodization. Hence, at the core of Koselleck's work is the attempt to replace the idea of linear, homogeneous time with a more complex, heterogeneous, and multilayered notion of temporality. In this essay I will demonstrate how this shift is achieved by means of three dichotomies: between natural and historical, extralinguistic and intralinguistic, and diachronic and synchronic time.