In this study, we investigate the impact factors on air pollution in terms of CO2, SO2 and NOx emissions simultaneously in China and compare changes in air pollution across provinces from 1998 to 2011 using a Multiple Indicators and Multiple Causes Model (MIMIC) within a Structural Equation Model (SEM) framework. Our findings reveal that GDP per capita and total population have the largest impacts on air pollution, followed by energy intensity, foreign direct investment, population density, and industrialization. The results also reveal that the inverted U-shaped Environmental Kuznets Curve (EKC) hypothesis exists in China. Our findings also demonstrate that Shandong, Jiangxi and Liaoning are the top three provinces with the most deteriorated air quality while Xinjiang, Fujian and Ningxia are with the best. These results not only contribute to advancing the existing literature, but also merit particular attention from policy-makers in China.