This paper has two related objectives. The first is to evaluate empirically whether annual data for China's GDP and its sectoral components from 1952 to 1998 can be modeled more accurately as a stationary process around a breaking trend function as opposed to a unit-root process. The second is to identify the long-run growth path of the Chinese economy and shocks that are big enough to have altered the path. The conclusion that China's major output time series are trend stationary with structural breaks has significant implications for the government in policy decisions for long-run growth and short-run stabilization. It also has implications for modeling comovements between output variables and other macroeconomic variables in cointegration analysis of the Chinese economy.J. Comp. Econom., December 2000, 28(4), pp. 814–827. Department of Commerce, Massey University (Albany), Auckland, New Zealand.