Romano-British urban origins have all too frequently been seen in terms of simple monocausal explanations, emphasising the role of purely military factors at the expense of the importance now attached to pre-Roman settlement nucleation. This article seeks to explore the question of small town origins and early development in the light of this wider perspective. It attempts to demonstrate that the period after A D 43 saw two overlapping and competing systems, one focused on pre-existing sites, the other on the new communications network, each with their own requirements and each with varying regional application and importance. Incorporation within the prevailing new order is shown to be more vital to urban development than preRoman or fort origins in isolation.'It can be shown that almost all the towns and small settlements occupied sites adjacent to the earlier forts and must thus have originated as civil settlements outside them, supplying some of the basic needs of the troops' (Webster 1966, 32).The search for 'origins' as a convenient explanatory anchor in the study of cultural development has been a common trend for numerous generations of archaeologists. This is as true of Roman Britain as of any other period. It is perhaps best illustrated by the monocausal explanations frequently advanced to cover questions of urban origins and development. Insufficient consideration has usually been given by Romanists to the potential importance of nucleation in the Iron Age, and a one-sided view has tended to predominate, explaining virtually all urban growth in Roman Britain as originating from purely military considerations. Such arguments, advanced by Webster (1966) in his important article on 'Fort and Town in Early Roman Britain', have been repeated and expanded for the cities by Rivet (1977) and for the small towns by Frere (1975). The latter article sought to demonstrate that some 56% of those small towns included in the survey had clear-cut or reasonable evidence for military origins, whilst only 13% (some 28% if one included sites with dual Iron Age/fort origin) had proven Iron Age predecessors. These apparently conclusive results have never been explored in detail, and this despite the importance now attached by prehistorians to pre-Roman settlement nucleation (Cunliffe 1976;1978; 1985). In his brief study of the small towns Todd (1970, 116) realised the complexity of the 0