Through a study of the history of the concepts of wealth and poverty, this paper investigates the onset of a transition in the conceptual architecture of epidemiological research concerning social differences in mortality rates from 1858 to 1914. It raises the question as to what the concepts of wealth and poverty meant to those who used them and what objects of interventions the conceptual architecture surrounding the concepts enabled the researchers to create. It argues that a transition began in the late 19th century in which an important framework for the understanding of causal relations behind the mortality patterns changed and that this change in turn influenced the scope of what was conceived as relevant objects of intervention.