Abstract: Based on the monitoring data from 1st July to 31th October, 2008 and 2009 at the Beijing urban ecosystem research station, correlations between PM_2.5 concentration and 6 meteorological factors were analyzed, including temperature, relative humidity, wind speed, vapour pressure, atmospheric pressure and wind direction. Main results showed that the dynamics of PM_2.5 concentration displayed an obvious fluctuation cycle every 6 weeks, while few changes happened within one week. The highest variation in weekly average of PM_2.5 concentration happened during the third 6 weeks, followed by the first 6 weeks, and the lowest variation occurred in the second 6 weeks. Correlation analysis suggested that the weekly average of PM_2.5 concentration was significantly correlated with all the 6 meteorological factors, and its correlation with the vapour pressure was the greatest. Results presented in this study confirmed that the weekly average of PM_2.5 concentration between July and August in Beijing could be estimated by the vapour pressure. The research would benefit the analysis and regulation of the pollution source of PM_2.5 in Beijing.