Based on the VAR model, this paper, adopting the data of China from 1981 to 2009, examines the long-term dynamic relationship among carbon emissions, energy consumption, economic growth and foreign trade by using the variable stationary test and co-integration test, impulse analysis and variance decomposition analysis. Results show that there is a long-time co-integration relationship among the four variables. Energy consumption and foreign trade have positive effect on carbon emissions, economic growth begins with a positive effect on carbon emissions, but at the later phase the former has a negative effect on the latter, and the extent of influence is sequenced by energy consumption, foreign trade and economic growth. Variance decomposition analysis indicates that foreign trade is the main variable to forecast mean square error of carbon emissions. Therefore, energy consumption is the least variable to forecast mean square error of carbon emissions.