BACKGROUND: The evidence between diurnal temperature range (DTR) and stroke remains controversial and sparse. We aimed to assess the relationship between DTR and emergency ambulance dispatches (EADs) due to stroke, and to explore whether there were effect modifications to the relationship. METHODS: A Quasi-Poisson generalized linear regression combined with a distributed lag non-linear model was used to examine the relationship between DTR and EADs for stroke between January 1st 2011 and June 30th 2018 in Guangzhou, China. We estimated the effects of the low DTR and high DTR (defined as DTR below and above 10 °C respectively) on EADs. The effects of minimum, maximum, 5th, 25th, 50th, 75th, and 95th percentiles of DTR compared with the DTR of 10 °C were also analyzed. RESULTS: A total of 20,275 EADs for stroke were included for analyses, among which 17,556 EADs were used in the model further adjusted for age and sex. A quasi-U-shaped relationship between DTR and EADs over lag0-2 days was observed. For the low DTR, per 1 °C decrease in DTR was significantly associated with an increase of 2.64% (RR = 1.03, 95% CI: 1.01-1.04) for EADs, while per 1 °C increase for the high DTR was non-significantly related with an increased risk of EADs (RR = 1.01, 95% CI: 0.90-1.13). Significant effects of the 5th and 25th percentiles of DTR on EADs were found when compared with the DTR of 10 °C. No significant effect modifications by age, sex or season were found to the association between DTR and EADs. CONCLUSIONS: We found a quasi-U-shaped relationship between DTR and EADs due to stroke in this study, while age, sex or season did not significantly modify the association between DTR and EADs. More high-quality evidence is needed to further explore and validate the relationship between DTR and stroke.