Abstract: On the basis of large and comprehensive field measuring data for atmospheric environmental parameters,a numerical forecasting model system on air quality was developed recently in order to make business forecast of air pollution in Jinan City.The system consists of pollution source model (SM),parameterization model on underlying source (XDM),meso-α meteorological model (M-αM),meso-β meteorological model (M-βM),parameterization model of turbulent statistics in the planetary boundary layer (PBLTM),dry and wet deposition model (DWDM),and high resolution chemical model(HRCM). The system was used successfully to do business numerical forecast of the air quality in Jinan.There is good agreement between forecasted and measured concentrations,and the average forecasting accuracy for 24 h can exceed 80%.