Building energy efficiency standards (BEES) are believed to be one of the most effective policies to reduce building energy consumption, especially in the case of the rapid urbanization content in China. However, there is little evidence backed up by measured data to validate the actual effectiveness of BEES in China. Using survey data collected from 1128 households in Chongqing China, this study applied the propensity scores matching method to estimate the effect of two BEES levels: the 50%-BEES (low level) and the 65%-BEES (high level). Results show that buildings built with 65%-BEES, on average, can reduce cooling and heating electricity use intensity (kWh/m2/a) by 41%, compared to buildings in the absence of the BEES. Meanwhile, the adoption of 50%-BEES can reduce cooling and heating electricity use intensity (kWh/m2/a) by 38%. However, energy savings are not significant if comparing buildings with 65%-BEES and 50%-BEES. The results indicate that there exists a performance gap between calculated design performance savings and actual operation energy savings. These empirical findings provide policymakers with valuable feedback on buildings' actual performance. The findings suggest that it is necessary to incorporate outcome-based compliance pathways into the current BEES system. Lastly, a data-driven building policy evaluation mechanism should be developed in China. Energy consumption databases should be developed to support policies such as building energy codes and standards' development and performance evaluation.