This paper models Chinese inflation using an output gap Phillips curve. Inflation modelling for the world’s sixth largest economy is a still under-researched topic. We estimate a partially forward-looking Phillips curve as well as traditional backward-looking Phillips curves. Using quarterly data from 1988 to 2002, we estimate a vertical long-run Phillips curve for China and show that the output gap, the exchange rate, and inflation expectations play important roles in explaining inflation. We adjust for structural change in the economy where possible and estimate regressions for rolling sample windows in order to test for and uncover gradual structural change. We evaluate a number of alternative output gap estimates and find that output gaps which are derived from production function estimations for the Chinese economy are of more use in estimating a Phillips curve than output gaps derived from simple statistical trends. Partially forward-looking Phillips curves provide a better fit than backward-looking ones. The identification of a non-increasing exchange rate effect on inflation during a period of large import growth hints at increased pricing to market behaviour by importers. This result is relevant to policies regarding possible exchange rate liberalisation in China.