Based on CO₂ emissions accounting method recommended by IPCC and energy statistical data, the CO₂ emissions of energy consumption in 30 provinces in mainland China were calculated. And the regression analysis has been done with the corrected DMSP/OLS night light data and CO₂ emissions under the same spatial unit. A 1km×1km grid CO₂ emissions of electricity consumption has been simulated and its spatial-temporal change in municipal scale also been analyzed. China's total CO₂ emissions of energy consumption in 2005, 2010 and 2013 was 5.702, 8.228 and 9.326 billion tons respectively. Accordingly, the CO₂ emissions of electricity consumption was 2.303, 3.562 and 4.207 billion tons respectively. The result shown that corrected DMSP/OLS night light data can be used to estimate CO₂ emissions much better than uncorrected data, and a relatively strong correlation between the total value of DN, statistic CO₂ emissions of energy consumption and electricity consumption existed at the provincial level. Relatively, total CO₂ emissions in developed areas is high but its intensity is low.