Recent research casts doubt on the commonly held notions that administrators affect student learning through instructional leadership and that changing administrators will improve school performance. To help construct a model for examining the process of leader succession that specifies a number of major school process and outcome variables associated with such succession, a literature review of succession factors based on two fundamental categories proposed by G. E. Gordon and N. Rosen (1981) is provided. The first of these categories, "Prearrival factors," focuses on reasons for succession, the selection process, and leader reputations and orientations. The second category, "Arrival factors," focuses on demography, school structures and processes, educational programs, successor actions, community environment, and school effectiveness. On the basis of literature reviews on each of these factors, it is argued that principal succession may have only marginal effects on subsequent school processes, structures, and outcomes. In conclusion, three longitudinally based empirical approaches to administrator succession--qualitative case studies, actuarial studies, and naturally occurring field experiments--are recommended and Gordon and Rosen's recommendation that leadership succession research replace more traditional leadership studies is affirmed. (JBM)