This research focuses on the spatial economics of geothermal district energy (DE) systems that distribute hot fluids through a pipeline network to multiple thermal customers. We argue here that DE is held back by uncertainty about its economic feasibility when implemented in real places. DE works best with high urban densities, but in the US the best geothermal resources are in the less-populated west. Economic geography theory suggests that the optimal size of a DE firm's service area should depend on the trade-off between the economies of scale inherent in a large service area and the greater network development costs and heat loss involved in serving a large area. The HEATMAP(C) software program developed by Washington State University was used to study the feasibility of numerous scenarios for a proposed DE system in Mammoth Lakes, California. Results indicate that the core area of the town is large enough, dense enough, and in close enough proximity to the resource for profitable DE development. Proposed ski village developments would increase the thermal load density and significantly enhance its ability to remain competitive with other fuels, even in pessimistic sensitivity analyses regarding geothermal heat input costs, public participation, and retro-fitting costs. (C) 2002 CNR. Published by Elsevier Science Ltd. All rights reserved.