In Sweden, where district heating accounts for a significant share of residential heating, it has been argued that improvements in end-use energy efficiency may be counter-productive since such measures reduce the potential of energy efficient combined heat and power production. In this paper we model how the potential trade-offs between energy supply and end-use technologies depend on climate policy and energy prices. The model optimizes a combination of energy efficiency measures, technologies and fuels for heat supply and district heating extensions over a 50 year period. We ask under what circumstances improved end-use efficiency may be cost-effective in buildings connected to district heating? The answer hinges on the available technologies for electricity production. In a scenario with no alternatives to basic condensing electricity production, high CO2 prices result in very high electricity prices, high profitability of combined heat and power production, and little incentive to reduce heat demand in buildings with district heating. In contrast, in a scenario where electricity production alternatives with low CO2 emissions are available, the electricity price will level out at high CO2 prices. This gives heat prices that increase with the CO2 price and make end-use efficiency cost-effective also in buildings with district heating. (c) 2013 Elsevier Ltd. All rights reserved.