The dynamic relationship between the PM2.5 and other kinds of air pollutants was analyzed based on vector auto-regressive(VAR) model. Methods of Granger causality test, generalized impulse response function and variance decomposition analysis were used, based on the daily data of the ambient air quality from January 1, 2013 to December 31, 2014 in Xi’an city to investigate the effect of other kinds of air pollutants on the PM2.5. The results show that the air quality system composed of PM2.5 and other kinds of air pollutants is stable, the increase of sulfur dioxide concentration, nitrogen dioxide and carbon monoxide will lead to the increase of PM2.5 concentration. Sulfur dioxide has the greatest effect on PM2.5 and the increase of ozone concentration will make PM2.5 concentration decrease.