Objective  Factors that affected the carbon emissions from land use in Guangdong Province were analyzed for improved environmental management.

  Methods   Based on the 1996-2017 statistics, a vector autoregression (VAR) model was constructed for the carbon emissions in Guangdong. Using Granger causality analysis, impulse response, and variance decomposition, the impact of land uses on the carbon emission in the province was quantitatively analyzed.

  Results   In the past two decades, the overall carbon emission from varied land uses in the province was on an upward trend. The per capita GDP, land use intensity, and energy intensity positively, but the structures of land use and industry negatively, correlated with the carbon emission. The contribution rates of those factor ranked by magnitude were land use intensity＞industrial structure＞energy intensity＞per capita GDP＞land use structure.

  Conclusion   It appeared that fostering transformation on energy structure and optimizing structure of land uses were the key maneuvers the province needs to consider in order to realize substantial reduction on carbon emission.