To quantitatively evaluate the effect of air pollution on hospital outpatient visits for respiratory diseases in Beijing. A semi-parametric generalized additive model (GAM) was used to analyze the exposure-effect relationship between air pollution and daily respiratory diseases outpatient visits from 2009 to 2011 in Beijing after controlling long time trend, the “day of the week” effect and confounding meteorological factors. At the same time, the model was established based on the sex and age groups. The results showed that there was certain lag effect of three kinds of air pollutants on daily respiratory diseases outpatient visits. The lag time of PM10 was 4-day moving average (avg03) or 6-day moving average (avg05). The lag time of SO2 and NO2 was 3-day moving average (avg02). An increase of 10μg/m3 in PM10, SO2, NO2 were significantly associated with the percentage increase (PI%) of 1.72%, 1.34% and 2.57% for all hospital outpatient visits, respectively. The elder more than 65 was the most sensitive to air pollution in Beijing, followed by those less than 14-year-old. The influence of air pollution on the female was more obvious than the male.