Abstract: When extreme temperature occurs in summer or winter, the number of stroke deaths increases significantly. Extreme temperature has different effects on stroke death in different regions. This study takes Ningbo as a representative coastal city and the data of stroke deaths and temperature in Ningbo from 2013 to 2019 are used. Using the distributed lag nonlinear model (DLNM), the quantitative lagged effects of extreme temperature on stroke death in different groups of people are studied. Results are as follows: (1) The effect of extreme high temperature is basically the same for all the groups except the young group. The cumulative relative risk (RR) and 95% confidence interval (95%CI) increases with time, indicating that the cumulative lagged effect of heat effect increases with time. It has a 3 d lagged effect on stroke death for all groups. The cumulative relative risk (95%CI) at 0—1 d, 0—2 d and 0—3 d lags at 40℃ are 1.29 (1.17—1.43), 1.38 (1.22—1.55) and 1.41 (1.25—1.60). (2) The effect of extreme low temperature has different trends in stroke death for different groups. For all the groups, there is no obvious lagged effect on the first day. There is an obvious lagged effect from the third to the fourth days. The cumulative relative risk and 95%CI increases with time, indicating that the cumulative lagged effect of cold effect increases with time. There is no significant lagged effect on stroke death after 15 days. The cumulative relative risk (95%CI) at 0—5 d, 0—10 d and 0—15 d lags at −4℃ are 1.23 (1.00—1.50), 1.61 (1.22—2.12) and 1.95 (1.39—2.75). (3) For different subgroups, extreme temperature poses higher risk on stroke death in the elder group (≥65 years old), and has no obvious effect in the younger group. Extreme high temperature leads to higher risk on stroke death for men than for women, while extreme low temperature poses higher risk on stroke death for women than for men.