Abstract: Based on the data from 2006 to 2016, the impact of various factors on direct and indirect carbon emissions in urban households in Beijing was studied using LMDI-I additive decomposition model. The research shows that, before 2013, the carbon emissions of urban households in Beijing were mainly from indirect carbon emissions, but the direct carbon emissions exceeded the indirect carbon emissions after 2013, which became the main source of carbon emissions for urban households' consumption. The increase in the living space per capita and the increase in consumption levels are the most important driving factors for the direct carbon emissions and indirect carbon emissions, respectively. The growth of urban population has a positive effect on both direct carbon emissions and indirect carbon emissions, while the reduction of energy intensity and energy conversion coefficient can effectively inhibit the growth of indirect carbon emissions.