OBJECTIVE To analyze the correlations between the incidence of allergic rhinitis (AR) and meteorological variables of previous periods, so as to establish non-linear prediction equations of AR in Beijing area. METHODS AR patients (10,478 cases) collected from Beijing Tongren Hospital during 2007-2010 and meteorological data (including daily average temperature, daily maximum temperature, daily minimum temperature, daily relative humidity, daily average vapor pressure, daily dew point temperature, daily precipitation, daily average wind speed, sea level pressure, and degree of comfort) collected from Beijing Municipal Meteorological Observatory in the same periods were used for the analysis. Non-linear correlation and regression were adopted to analyze the relationship between AR incidence and meteorological variables of former six-qi stage which was defined according to Yunqi theory of Chinese medicine. Comprehensive meteorological parameter was introduced to establish the predictive model. RESULTS The high incidence of AR appeared in the 4th qi stage (from the Beginning of Autumn to Autumn Equinox), while the changes of meteorological variables appeared in the 3rd qi stage (from Grain in Beard to Greater Heat), which advanced one phase. The incidence of AR was closely associated with vapor pressure. The correlation coeffifi cients of two predictive models were between 0.8931-0.9176 and all of them have passed signififi cant statistical tests, which showed a satisfactory fifi tting effect. CONCLUSIONS Comprehensive meteorological parameters can be used to forecast AR incidence, which is benefifi cial to AR prevention.