Infrastructure costs, lot size, and developers' response: the need for an assessment Planners continue to be concerned about large-lot residential development in exurbia (Nelson and Sanchez, 2005). (1) A substantial literature has attempted to explain why this form of development persists [see Byun and Esparza (2005), for a recent synthesis].Missing in the discussions, however, is in-depth research on how on-site costs for infrastructures may influence the decisions of single-family residential land developers (hereafter called developers) to produce large-lot exurban development rather than small-lot development. The omission is surprising, given that developers are integral to the development process (Peiser, 1990) and that their responses to regulations have been seen as contributing to sprawl (Byun and Esparza, 2005;Dowall, 1984;Frieden, 1979). The oversight has resulted in a lack of knowledge about how developers view developmentöin particular, why relatively large-lot exurban development remains popular with developers despite policies that attempt to steer them elsewhere.The objective of this paper is to further our understanding about why developers might be attracted to large lots in exurbia. I do so by examining how developers' on-site costs per lot for the basic infrastructures of sewer and water services and roads vary with lot size, and in turn, how these costs can affect the lot sizes that developers choose. It is not the intention of this paper to address the many other factors that influence the choice of lot sizes historically observed in land development.I utilize on-site costs as the basis for analyses because this measure highlights a conundrum in the literature which, if resolved, can help us to understand why developers choose larger over smaller lots. The conundrum lies in contradictory findings in the literature on the costs of sprawl, which has concluded that on-site costs per lot increase as lot sizes increase [see Najafi et al (2007) and Speir and Stephenson (2002) for the latest findings in this regard], and the literature on developer decision making,