The current study was based on annual ambient air quality monitoring data and corresponding meteorological observation data of Xi'an in 2011. Distribution models on hourly concentrations of PM10, SO2, and NO2 were studied, and the results showed that statistical distribution functions varied from seasons and from pollutants. The optimal distribution models of PM10 concentrations in the four seasons (spring, summer, autumn, and winter) were generalized extreme value distribution (GEVD), Weibull, Weibull, and GEVD, respectively; those of SO2 were lognormal, log-logistic, log-logistic, and GEVD, respectively; and those of NO2 were Weibull, lognormal, GEVD, and GEVD, respectively. The concentrations ranges were 0.03 ∼ 0.20 mg/m3 for PM10, 0.008 ∼ 0.17 mg/m3 for SO2, and 0.01 ∼ 0.12 mg/m3 for NO2, and the probabilities of concentrations in the ranges for accordingly pollutants were up to 85 %. Effects of the meteorological parameters on concentrations of PM10, SO2, and NO2 were studied with linear correlation analysis method for Xi'an city. The results indicated that pollutant concentrations had a negative correlation with wind speed, temperature, and mixing height (MH), whereas it had a positive correlation with atmospheric pressure and atmospheric stability. Both temperature and atmospheric pressure were the most obvious correlation with SO2 concentration with r value of −0.7916 and 0.7032, respectively. Wind speed and MH had the most obvious correlation with NO2 concentration with r value of −0.4423 and −0.3997, respectively. SO2 had the best correlation with meteorological parameters. Analyzing the statistical characteristics of urban air pollution concentration and their relationships with meteorological parameters are of great importance to study urban air pollution problems and corresponding prevention measure.