Taking the debate on evidence-based management and success-factor research as a point of departure, this paper discusses three different views of the role of identified regularities for management practice: (1) acceptance of regularities, where regularities are understood as delimiting the scope of action, (2) use of regularities, where regularities are understood as restrictions that change over time and (3) management of regularities, where regularities are understood as temporary invariances that can be managed by the firm. Reflecting on the scientific status of regularities in the social sciences, we argue that the first two views are based on a limited understanding of the nature of management sciences.