OBJECTIVES: Short-term temperature variability (TV) is associated with the exacerbation of asthma, but little is known about the relative effects of intra- and inter-day TV. We aimed to assess the relative impacts of intra- and inter-day TV on childhood asthma and to explore the modification effects by season. METHODS: A quasi-Poisson generalized linear regression model combined with a distributed lag nonlinear model was adopted to evaluate the nonlinear and lagged effects of TV on childhood asthma in Shanghai from 2009 to 2017. Intra- and inter-day TV was measured with diurnal temperature range (DTR) and temperature changes between neighboring days (TCN), respectively. RESULTS: Increased DTR was associated with the elevated relative risk (RR) of daily outpatient visits for childhood asthma (DOVCA) in both the whole year (RR(lag0-14) for the 99th percentile: 1.264, 95% confidence interval (CI): 1.052, 1.518) and cold season (RR(lag0-12) for the 99th percentile: 1.411, 95% CI: 1.053, 1.889). Higher TCN in the warm season was associated with the increased RR of DOVCA (RR(lag0-14) for the 99th percentile: 2.964, 95% CI: 1.636, 5.373). The number and fraction of DOVCA attributed to an interquartile range (IQR) increase of TCN were higher than those attributed to DTR in both the whole year period and warm season. However, the number and fraction of DOVCA attributed to an IQR increase of DTR were greater than those attributed to TCN in the cold season. CONCLUSIONS: Our results provide novel evidence that both intra- and inter-day TV might be a trigger of childhood asthma. Higher DTR appeared to have greater impacts on childhood asthma in the cold season while an increase in TCN seemed to have bigger effects in the warm season.