Abstract: Air pollution is not only associated with emissions of pollutant sources, but also has close links with meteorological factors such as temperature and precipitation. Establishment of the relationship between regional meteorological data and air quality is of great importance of the study of air quality and its changes. Because of the non-linear relationship between air pollution and meteorological conditions, the meteorological data and the actual concentration of air pollutants 20 days before as an input parameter dynamically created the daily multiple linear regression equation, then substituted the forecast meteorological data with the obtained equation to solve the predictive values and air pollution index (API) values of SO₂, NO₂, and PM₁₀. The preliminary test prediction demonstrated that the sliding forecast method of 20 days as a time window could more accurately forecast the air quality of the Nyingchi and had certain application and promotion values. By using this method, meteorologists took adequately into account the complex and dynamic relationship between air pollution and meteorological conditions, thus overcoming the shortcomings of the traditional static air quality forecasting method.