Information asymmetry between governments and industries is a major impediment to set an optimal energy efficiency target for reduction in energy demand. Typically, the regulators face challenges of incomplete and imperfect information on the minimum improvement in energy intensity that industries would achieve in baseline cases and the maximum technical potential of energy efficiency that industries are able to achieve. By applying a two-tier stochastic frontier model based on provincial panel data 2006–2015 in China, we disentangle the impacts of information asymmetry between the governments and firms on industrial energy intensity from other major regulatory, technological and economic factors. Results show that, on average, nearly 75% of the unexplained variation in industrial energy intensity is attributed to the information gap between firms and governments. On average, the realized energy intensity is upper biased by 7.25% due to the information gap compared to the expected level without the information gap although the net impacts vary considerably across provinces and over time. One observation is that firms have much stronger impact on the energy intensity in the period when the government prioritized economic growth rather than energy savings. In addition, the impact of firms and governments varies markedly across the quartiles of economic development level, technological progress, policy stringency, and energy resource endowment.