Global cities produce more than 70% of the world's CO2 emissions and thus play an important role in addressing climate change. Few statistics are available with respect to national city-level energy consumption in China-as such, in this study we apply spatiotemporal modeling in order to assess China's city-level CO2 emission levels using DMSP/OLS nighttime light imagery. We examine the spatiotemporal variations and determinants of CO2 emissions using a series of distribution dynamic approaches and panel data models for proposing feasible mitigation policies, the results of which show that while per capital CO2 emissions were characterized by significant regional inequalities and self-reinforcing agglomeration during the study period, regional disparities decreased and spatial agglomeration gradually increased between 1992 and 2013. The results of our estimation further reveal the importance of economic development, population growth, industrial structure, and capital investment as the factors positively affecting China's city-level per capita CO2 emissions. Conversely, FDI is found to exert a negative impact. Our results also strongly support the existence of an inverted U-shaped relationship between per capita CO2 emissions and economic development, thereby confirming the EKC hypothesis. The findings obtained in this study could provide important decision-making support in the task of building China's low-carbon cities of the future. (C) 2017 Elsevier Ltd. All rights reserved.