Scholarly understanding of the origins of the bipartite manorial economy of the early middle ages has long been vitiated by our relative ignorance of the agrarian relations of production that characterised the late Roman period. In recent years, however, our understanding of the agrarian economy of the late Roman empire has advanced considerably. On the basis of the legal, papyrological, and archaeological evidence, it would appear that from the fourth century onwards, highly commercialised bipartite estates, owned by members of an emergent aristocracy of service became a common feature of life in the Roman world. These estates are at their most visible in the evidence from the eastern provinces of the empire, but they can be identified in the west as well. Indeed, it is argued that such estates can be traced in the west from the fifth century through to the eighth. In terms of economic structure, therefore, the bipartite estates of the early middle ages would appear to have been a late Roman survival. In post-Roman conditions, however, and especially as a result of demonetization and deurbanization, these estates took on an increasingly autarkic aspect. Production for consumption and use supplanted production for exchange.