This paper takes a distributed perspective on leadership practice in schools. Drawing upon empirical evidence from two contemporary studies of successful school leadership and the international research base concerning distributed leadership, the paper argues that leadership practice is a distributed entity that is mediated through human interaction. It suggests that leadership activity at the level of the school, rather than the level of the individual, is the appropriate unit of analysis in studying leadership practice. The paper argues that the distributed perspective offers a new and important conceptual lens through which leadership practice in school can be understood and analyzed. It concludes that further research is required to investigate the interface between distributed leadership and school improvement to ensure that this particular perspective on school leadership is not misleading. (Contains 40 references.) (Author)