Few epidemiological studies have been reported as to whether there was any interactive effect between temperature and humidity on respiratory morbidity, especially in Asian countries. The present study used time-series analysis to explore the modification effects of humidity on the association between temperature and emergency room (ER) visits for respiratory, upper respiratory tract infection (URI), pneumonia, and bronchitis in Beijing between 2009 and 2011. Results showed that an obvious joint effect of temperature and humidity was revealed on ER visits for respiratory, URI, pneumonia, and bronchitis. Below temperature threshold, the temperature effect was stronger in low humidity level and presented a trend fall with humidity level increase. The effect estimates per 1 °C increase in temperature in low humidity level were -2.88 % (95 % confidence interval (CI) -3.08, -2.67) for all respiratory, -3.24 % (-3.59, -2.88) for URI, -1.48 % (-1.93, -1.03) for pneumonia, and -3.79 % (-4.37, -3.21) for bronchitis ER visits, respectively. However, above temperature threshold, temperature effect was greater in high humidity level and trending upward with humidity level increasing. In high humidity level, a 1 °C increase in temperature, the effect estimates were 1.84 % (1.55, 2.13) for all respiratory, 1.76 % (1.41, 2.11) for URI, and 7.48 % (4.41, 10.65) for bronchitis ER visits. But, there was no statistically significant for pneumonia. This suggests that the modifying effects of the humidity should be considered when analyzing health impacts of temperature.