Diurnal temperature range (DTR) is a critical indicator that reflects the intra-day variability of temperature. In this study, we used Poisson regression with a distributed lag non-linear model (DLNM) to fit the exposure-response relationship between DTR and cardiovascular disease (CVD) hospital admissions and assess the lag effect simultaneously. The CVD and meteorological data were collected from January 1, 2014 to December 31, 2015. A non-linear relationship was found between DTR and CVD hospital admissions, and DTR had a prolonged lag period of up to 21 days. The impact of low DTR was significantly more deleterious than that of high DTR. Stratified analysis revealed that male and the elderly people (≥ 65 years old) were more vulnerable to low DTR exposure than female and adult (<65 years old), and there were no significant differences in both gender and age subgroups for high DTR exposure. Thus, evaluating the impact of DTR on CVD hospital admissions could help in the development of effective preventive measures for local public health departments.