The relation between coal consumption and the economic growth during 1952—2006 in China was tested by Bai and Perron's multiple structural breaks approach.  Results show that the two series are segment-trend-stationary process around two or three structural breaks.  After detrending, The VAR results indicate that in the sample period there is only one-way causality between the coal consumption and GDP. The conclusion is useful to formulate the energy policies as well as causality among the other economic series.