Facility location decisions require large investments of capital and represent long-term commitments for businesses. For firms that are highly dependent on access to raw materials, plant location has a major impact on direct material costs and, hence, on total product costs, pricing, and profitability. In the procurement of bulky, low-value raw materials, the cost of transportation represents a large portion of the final delivered cost. Depending on the commodity, the price paid is often a function of the maximum delivered cost (marginal price) of attaining a target quantity at a particular location. Such costs differ considerably from one candidate location to another if there is significant spatial variation in the quantities of the raw materials and the costs to produce them. In addition, the quality of a transportation network can play a role in the relative marginal prices among candidate locations. An adequate assessment of marginal price is essential for identifying promising areas for plant location. In this paper, we describe a methodology for generating marginal price surfaces when the number of potential supply points and the number of candidate plant locations are both very large. We present a GIS-based application of the methodology for the identification of promising locations for switchgrass-to-ethanol conversion plants.