The effects of anthropogenic emission inventories on simulated air pollutants concentrations in Zheijiang Province have been analyzed using the WRF-Chem (Weather Research Forecast-Chemistry) model. Three independent emission inventories, i.e. INTEX-B (Intercontinental Chemical Transport Experiment-Phase B), REASv2.1 (Regional Emission Inventory in Asia version 2.1), and HTAP_v2 (Hemispheric Transport of Air Pollution version 2), are used for model simulations during December 2013. The three experiments are denoted by IN, RE, and HT, respectively. Compared with in situ measurements, the three experiments can reasonably reproduce the temporal and spatial characteristics of PM2.5, PM10, and NO₂ surface concentrations with correlation coefficients ranging from 0.5 to 0.8. More than 85% of simulated values are within the range of 0.5 to 2 times of observational values. However, all of them have a poor performance on simulation of SO₂ concentration. The relative biases of PM2.5 and PM10 concentrations simulated by IN, RE, and HT are about 30%, 16%, and 6%, respectively, and the best performance is obtained by HT. The PM2.5 primary emissions and the secondary aerosol precursor SO₂ emissions of INTEX-B are significantly higher than those of REASv2.1 and HTAP_v2 emission inventories, which results in more sulfate aerosols and subsequently increases the PM2.5 concentration. The obviously lower NH₃ emission of HTAP_v2 compared to that in the other two emission inventories inhibits the formation of nitrate aerosols, which helps to reduce the PM2.5 concentration. Differences between the base year of emission inventories and the simulation year have greater impacts on the accuracy of simulated SO₂ concentrations than that of PM2.5, PM10, and NO_x. SO₂ emissions of INTEX-B are significantly higher than those of REASv2.1 and HTAP_v2 emission inventories, especially for the northern part of Zhejiang Province and the coastal industrialized areas, which is the primary reason for the obvious overestimation of SO₂ using the INTEX-B inventory. NO_x emissions of the three emission inventories are very consistent over Zhejiang Province, which could be the main reason for the similar modeled values and small relative biases (-8%-4%) of NO₂ in the three experiments. Furthermore, the simulations with different anthropogenic emission inventories do differ in their predictions of daily PM2.5, PM10, SO₂, and NO₂ concentrations with mean variations of 14%, 15%, 51%, and 16%, and maximum variations of 69%, 78%, 137%, and 132% over Zhejiang Province. The variations of monthly average concentrations of pollutants are basically consistent with those of daily average values of pollutants, but the maximum variations of simulated monthly values are obviously lower than those of simulated daily values. Generally, the variability of PM2.5, PM10, and NO₂ in the three simulations is significantly smaller than that of SO₂.