Using the quasi-natural experiment that adjusts the levy standards of sulfur dioxide (SO2) emissions in different Chinese provinces between 2007 and 2013, this paper examines the effect that such levy standard change exerts on pollution control by adopting the difference in differences (DID) model. Although the effect of the pollution charging policy in China is a widely debated issue due to the low per-unit charge and an “endogenous enforcement of the law”, this paper finds that the increase in pollution charges significantly reduced firms’ SO2 emissions by 27%. This finding remains robust after evaluation with the parallel trend test, placebo tests and other robustness check methods. Compared with state-owned enterprises (SOEs), non-state-owned enterprises are more likely to be driven by policies. In discovering influence mechanism, this paper finds that the increase of levy standard urged firms to cut back consumption of raw coal, fuel coal and heavy oil, while the consumption of clean gas has no significant increase. Only SOEs purchased and upgraded their desulfurization equipment after the reform. This paper also furthers discussion on the externality of the reform. This paper finds no evidence for pollution transferring, but confirms the reduction effect on air SO2 concentration from the levy standard reform.