A New Du Boisian Sociology has recently clarified, elevated, and synthesized Du Bois’s sociological contributions. We argue that more systematic and detailed study of Du Bois’s research methodologies, with an eye towards their contemporary applicability, can further strengthen this body of scholarship. Here we begin this effort with sustained attention to Du Bois’s use of quantitative data and methods during a productive and illustrative period around the turn of the twentieth century (1898–1902). This adds a level of depth and specificity to a subset of existing Du Bois scholarship that has more generally noted quantitative inquiry as one aspect of Du Bois’s social-scientific approach of mixed-methods triangulation. We detail how and why Du Bois developed an inductive, theoretically generative approach to his research on race. This orientation appears, at first glance, to be a misfit for contemporary quantitative sociology, which is currently skewed towards deductive theory testing and causal inference. We demonstrate that Du Bois’s quantitative methodology invites sociologists to return to exploratory, descriptive, and theoretically generative quantitative research based on creative syntheses of primary and secondary data that span generations and levels of institutional and geographic aggregation. Such data can, among other possibilities, assess within- and between-race comparisons and intersections of race with factors including class, gender, age, place, and time. Our study also enters Du Bois, as historical precedent, into current debates regarding quantification’s productive role, if any, in social science research on race/racism and other axes of systemic inequality.