Aiming at the problem about factors affecting the content of PM2.5, Using collected data about January 1, 2015 to July 31, 2015 daily AQI index and the six basic monitoring indicators in Bengbu，conducting quantitative research about the relationship between PM2.5 content and other five sub-indicators and their corresponding levels of contaminants by applying multiple iinear regression model,and through the goodness of fit test, F test, t test method to test the model , and finally get accurate and reliable multiple linear regression model ,this model has a high degree of fit, simple and intuitive, and other advantages, providing a strong reference for the multiple linear regression model in PM2.5 content analysis.