Rapid urbanization has significant effects on China’s CO₂ emissions and contributes to climate change. It is of great theoretical and practical significance to explore the influencing factors of carbon emissions. Both the demographic composition and residential consumption of urban-rural residents were included in the expanding vector form of Kaya identity. This study employed the Logarithmic Mean Divisia Index method (LMDI) to investigate the impact of urbanization on CO₂ emissions from the perspective of residential consumption in China during 1980-2020. Six effects were included population size, urbanization, residential consumption, consumption restraint, energy intensity and emission coefficient. The results showed that residential consumption contributed far more to carbon emissions than the other five effects. The impact of urbanization on carbon emission had been greater than that of population size since the middle 1990s. The decline of residential consumption rate imposed stimulation rather than restraint on carbon emissions due to the synchronous growth of shares of higher carbon investment and export in the economy. The continuous decline of energy intensity has an obvious effect on carbon emission reduction. The research results are expected to give inspiration to guide the direction how to optimize resident consumption and to achieve green and low-carbon urbanization development.