This paper examines empirically a set of propositions on organizational stratification drawn from functional theory. The data consist of 38 departments of a large university which are treated as contextual units, as well as the responses of individual faculty to a survey. The findings present an interesting theoretical paradox: (1) functional arguments do not appear adequate to account for differences in income stratification among departments, but (2) they do explain participants' beliefs and judgments about the legitimacy of organizational reward systems. The implications are discussed, and a line of inquiry is suggested that would focus on the problem of establishing legitimacy with the environment(s) as a major source of organizational structure. It is suggested that organizational features act as symbolic devices that inform others that a given organization is a competent member of its set. In short, it is conforming to agreed upon definitions of the appropriate—most “rational”—structure for thattype of organization.