Increasing studies have reported temperature modification effects on air pollutants-induced respiratory diseases. In the current study, daily data of respiratory emergency room visits (ERVs), meteorological factors, and concentrations of air pollutants were collected from 2013 to 2016 in Lanzhou, a northwest city in China. Daily average temperature was stratified into low (≤ 25 percentile, P25), medium (25-75 percentile, P25-P75) and high (≥ 75 percentile, P75) to explore how temperature modifies the effects of air pollutants (PM2.5, PM10, SO2, and NO2) on respiratory ERVs by using generalized additive Poisson regression model (GAM). Seasonal modification was also investigated. Results showed that (a) PM10, PM2.5, and NO2 had the strongest effects on respiratory ERVs in low temperature; (b) males and 15-and-younger were more vulnerable in low temperature while females and those older than 46 years were highly affected in high temperature; (c) PM10, PM2.5, and NO2 were mostly associated with the total and both males and females in winter, while SO2 resulted in the highest risk for the total and males in autumn and females in spring. In conclusion, this study found significant temperature modification effects and seasonal differences on the risks of respiratory ERVs due to air pollutants in Lanzhou, China.