This study focuses on the role played by emissions from transportation, industry and power generation on the concentrations of O 3 , CO, NO, NO y and SO 2 in Guangdong province of China. Observational data of the pollutants and numerical modeling of atmospheric chemistry, transport and removal processes with STEM-2K1 and MM5 are used for March 2001. The objective is to identify the relative importance of the three emission sources on the concentrations of the pollutants. In addition, the relative importance of NO x and VOC emissions from the transportation sector for O 3 production is examined. The observations at a rural and an urban site in the region show distinctive characteristics, indicating the importance of local emissions. A comparison of the observed ratios of CO/NO x and SO 2 /NO x at the two sites with those derived from the emission inventories show the usefulness of the emission inventories. A control simulation with all emissions included shows a fairly satisfactory performance of the model in terms of the comparison between the observed and modeled concentrations of CO, SO 2 , NO y and O 3 at the rural site, although the observed extremes are less well simulated. Three simulations with different emission scenarios suggest that the transportation source was the main contributor to NO x , CO and O 3 concentrations accounting for 34.2%, 33.1% and 17.8% of their total concentrations, respectively. For SO 2 concentration, the main contributor is the power plant sources with a contribution rate of 32.9%. For SO 2 , the doubling of emissions from the industry has a much larger effect on the ambient levels than the doubling from the transportation, whereas the increase in the transportation sector sources has approximately twice the effect of the increased industrial emission on O 3 . The doubling of emissions from the transportation and industry sources would increase the O 3 concentration by about 26.7% and 8.5% for the whole region. Finally, the O 3 /NO z ratio of 25 was found to be a reasonable threshold to separate NO x -limited and VOC-limited regimes concerning O 3 chemistry. It is found that the urban area is VOC-limited and the non-urban area is NO x -limited. These results have implications for O 3 control strategy for the region.