This study analyzed particulate matter number concentration spectral distribution,variation features of particulate matter mass concentration and their relationships to weather conditions using concentration data of particulate matter from monitoring instruments Grimm180,the number concentration data of PM_(10),PM_(2.5),PM_(1.0) and routine weather data observed in the air component monitoring station in Anshan in 2007.The results showed that the number concentration spectrum agreed well with the results of Junge distribution.There was a significant negative correlation between υ value and visibility.When υ value was greater and the content of coarse particles within small particles was less than 90%,the visibility was poor.There were two peaks appearing in the daily change of particulate matter mass concentration.The mass concentrations of PM_(10),PM_(2.5) and PM_(1.0) had good correlations;PM_(2.5)/PM_(10) had an average of 0.654;PM_(1.0)/PM_(2.5) had an average of 0.832;and PM_(1.0)/PM_(10) had an average of 0.545.Because the dominant wind direction in Anshan is SE,the mass concentration change of particulate matter is affected little by the Liaoning dust mobile path,and mainly affected by the cumulative emission of pollutants.The mass concentration of particulate matter in fog weather is higher.The occurrence frequency of fog weather and the mass concentration change of particulate matter have positive correlation.During fog weather,the intercept of the regression equation is bigger than the average annual return of the equation,which has a direct effect on the study of mutation of particulate matter mass concentration.