Two Gaussian air quality dispersion models, the industrial source complex short-term model (ISCST3) with and without modification have been used to simulate the pollutant concentration distribution in urban areas based on the meteorological data and the emissions distribution of sulfur dioxide. The verified data show that the modified model is more accurate in the urban area of Shijiazhuang. Using the modified model predictions, the control strategies of sulfur dioxide in the urban area have been studied, and the result show that the second long-term (to 2010) strategy can mitigate air pollution significantly and maintain pollution levels within permissible limits.