When planning a neighborhood-scale distributed energy system (NDES), enough attention should be paid to the urban spatial structure, which may affect not only the system configuration on the supply side but also the load profiles on the demand side as well as their interactions. In this study, a novel nonlinear optimization model is developed for the determination of neighborhood design and NDES arrangement in terms of energy-saving aspect. By executing the optimization model, besides the configuration of energy supply technology, optimal building mix within the neighborhood can be also deduced. As an illustrative example, a mixed-use neighborhood located in Shanghai, China has been assumed for analysis. According to the simulation results, the introduction of NDES may achieve satisfied energy-saving benefits. Although additional energy consumption including heat loss and pump power may be encountered in the NDES, it can be offset through rational plan of the urban spatial structure and its cooperation with the supply side. For a fixed building mix, the reduction of supply radius may lead to better energy performance. Moreover, there exists the maximum supply radius over which, the NDES will not be recommended from the energy-saving viewpoint. (C) 2017 Elsevier Ltd. All rights reserved.