Cellulosic ethanol has been identified as a promising alternative to fossil fuels to provide energy for the transportation sector. One of the obstacles cellulosic ethanol must overcome in order to contribute to transportation energy demand is the infrastructure required to produce and distribute the fuel. Given a nascent cellulosic ethanol industry, locating cellulosic ethanol refineries and creating the accompanying infrastructure is essentially a greenfield problem that may benefit greatly from quantitative analysis. This study models cellulosic ethanol infrastructure investment using a mixed integer program (MIP) that locates ethanol refineries and connects these refineries to the biomass supplies and ethanol demands in a way that minimizes the total cost. For the single- and multi-state regions examined in this study, larger facilities can decrease ethanol costs by $0.20-0.30 per gallon, and placing these facilities in locations that minimize feedstock and product transportation costs can decrease ethanol costs by up to $0.25 per gallon compared to uninformed placement that could result from influences such as local subsidies to encourage economic development. To best benefit society, policies should allow for incentives that encourage these low-cost production scenarios and avoid politically motivated siting of plants. (C) 2010 Elsevier Ltd. All rights reserved.