Over the last 150 years, visualizing data has been a core issue of making and communicating science. Visual representations of data have not only shaped our understanding of the social world, but they also have contributed to the “scientification” of social science itself. By actively promoting the application of the “graphical method” to statistics, early statisticians opened a new path for making social facts visible. Interestingly, these strong advocates of “statistical images” fostered the increasing legitimacy of imaging quantitative information, while also striving mightily to give a concrete and tangible shape to allegedly abstract knowledge, fostering therefore a sensory approach to statistics. This article examines the primary importance of thinking the materiality of visual statistics. Overriding the long-standing distinction between objectivity and subjectivity, and between visuality and materiality, it argues that a greater attention should be paid to the appropriation process of statistical knowledge at a critical moment of its institutionalization and acknowledgement as a science.