This study examines the effect of industrial land market reform on the local governments' land supply decision in China. Since 2007, the central government has required that each land parcel for sale must be sold to a proposed industry through public auction. The unit transaction price of these sales (yuan per square meter) must be higher than the county-level minimum price, unless permitted by the central government. Therefore, the local governments are no longer able to sell industrial land parcels at low prices to attract firms and are forced to consider the comparative advantages of their land for particular industries. In order to analyze the local governments' land supply decision after the reform, I develop a theoretical model in which I treat the local governments as monopolist land suppliers subject to the central government's auction and minimum price restrictions. The model predicts that the local governments are responsive to the comparative advantages. In addition, when the relationship between the transaction price and the minimum price changes, the local governments can be more or less responsive to their comparative advantages, depending on the revenue from land sale, the positive externalities from the firms, and the reference of the central government. I employ detailed parcel-, county-, and industry- level data to test the model predictions. My empirical findings suggest that a one percentage point increase in the employment share of an industry leads to a 1.3% - 1.6% increase in the land supplied to that industry ceteris paribus. Moreover, the local governments become more responsive to the comparative advantages when the transaction price is lower than the minimum price. These findings verify the implications of the theoretical model. Furthermore, the effects are heterogeneous among counties; the local governments of more developed counties are more responsive to the comparative advantages.