Early archaeological research on the Iron Age of Uganda focused upon earthworks, such as Bigo, with the purpose of validating historians' interpretations of oral traditions. Recent research has continued the emphasis upon large sites but with archaeological interpretation given precedence over historical reconstructions. This paper discusses archaeological surveys undertaken in western Uganda in 1991 with the aim of examining Iron Age settlement patterns from a regional perspective, in which the large sites form only one element. Pottery analyses permit the establishment of a tentative chronology, which in combination with data on site sizes facilitates a new perspective on state formatioh in the region. The proposed model of the development of social complexity is compatible with revisionist interpretations of the historical evidence. Together, they suggest that the Nyoro state emerged after several centuries characterised by competing and often unstable small polities or chiefdoms.