The process of energy conservation and emission reduction in China requires the specific and accurate evaluation of the energy efficiency of the industry sector because this sector accounts for 70 percent of China's total energy consumption. Previous studies have used a "black box" DEA (data envelopment analysis) model to obtain the energy efficiency without considering the inner structure of the industry sector. However, differences in the properties of energy utilization (final consumption or intermediate conversion) in different industry departments may lead to bias in energy efficiency measures under such "black box" evaluation structures. Using the network DEA model and efficiency decomposition technique, this study proposes an adjusted energy efficiency evaluation model that can characterize the inner structure and associated energy utilization properties of the industry sector so as to avoid evaluation bias. By separating the energy-producing department and energy-consuming department, this adjusted evaluation model was then applied to evaluate the energy efficiency of China's provincial industry sector. (C) 2015 Elsevier Ltd. All rights reserved.