To find the key factors of the carbon emissions from energy consumption in three northeastern provinces, a hybrid approach was applied by combining the extended Kaya identity method and the logarithmic mean Divisia index (LMDI) decomposition method. Firstly, an optimized carbon emission decomposition model was developed to measure and decompose the carbon emissions and carbon emission intensity of three northeastern provinces during 2005~2016. Then, by performing a comparative analysis of three northeast provinces and the whole China, the impacts of the following factors on the carbon emissions were examined:the energy structure, the energy intensity, the industrial structure, the economic output and the population. The empirical results showed that, during 2005~2016, the total carbon emissions of three northeastern provinces accounted for 8.84% of the total carbon emissions of China, and the average carbon emission intensity of these three provinces was higher than that of China. Moreover, both economic output and population had significant positive effects on the carbon emissions; meanwhile, the largest contribution of economic reached 188%. The economic development and urbanization did not reduce the carbon emissions in these three provinces. In addition, the industrial energy intensity, energy structure and industrial structure had significant negative effects on carbon emission in which the maximum effect of energy intensity reached 59%. It was also found that there exists a large adjustment space in the industrial energy intensity. In summary, to promote the development of low-carbon economy, we suggested to reduce energy consumption, adjust the internal structure of the industry and improve the economic policy system in three northeastern provinces.