Abstract: In this paper, we use the methods of unit root test, cointergration test, Granger causality test and variance decomposition to empirically study the dynamic relationship of energy consumption, carbon emission and economic growth based on statistical data from 1995 to 2014 in China. The main conclusions from the study are that: there is a cointergration relationship among energy consumption, carbon emission and economic growth; cointergration relationship formula shows that energy consumption has more effect on carbon emission than economic growth; the results of Granger causality test find that there is a two-way Granger causal relation between carbon emission and energy consumption; energy consumption and carbon emission are the one-way Granger cause of economic growth; the results of variance decomposition suggest that carbon emission has significant influence on energy consumption and economic growth.