Extreme temperature has been reported to be associated with an increase in acute disease incidence in several cities. However, few similar studies were carried out in Shenzhen, which is a subtropical city located in the southern China. This study explored the relationship between the emergency incidences and extreme temperatures, and investigated the role of air pollutants played in the temperature-related effects on human health in Shenzhen. We conducted a distributed lag nonlinear model study on the effect of extreme temperatures on emergency incidences in Shenzhen city during 2013–2017. Here, only the total emergency incidences, emergency incidences for respiratory diseases, and cardiovascular diseases were taken into consideration. Air pollution, subgroups, and seasons were adjusted to investigate the impacts of extreme temperatures on emergency incidences. Relative risk (RR) and 95% confidence intervals were calculated with the R software. From lag 0 to 21 days, the RR of temperature-total emergency department visits, temperature-cardiovascular, and temperature-respiratory diseases was 1.09 (95% CI: 0.98–1.20), 1.22 (95% CI: 0.96–1.56), and 1.06 (95% CI: 0.70–1.60) at extremely low temperature (first percent of temperature, 10 °C), respectively. During the same lag days, the RR was 1.02 (95 % CI: 0.92–1.14), 0.64 (95% CI: 0.49–0.86), and 0.92 (95% CI: 0.56–1.53) between extremely high temperature and total emergency department visits, cardiovascular, and respiratory diseases, respectively. The cumulative effects gradually went up with time for all types of emergency incidences in warm seasons (5 days moving average of temperature < 22 °C). However, the cumulative effects of total emergency incidences and Cvd emergency incidences were increased within the first lag 5 days, and then decreased until lag 21 in hot seasons (5 days moving average of temperature ≥ 22 °C). The cumulative effects of Res emergency incidences showed a declined trend from lag 0 to lag 21. The elderly (≥ 65, P1: RR = 1.49, 95% CI (1.30, 1.71); P99: RR = 0.86, 95% CI (0.71, 1.04)) and men (P1: RR = 1.27, 95% CI (1.14, 1.42)) seemed to be more vulnerable to extreme temperature than the younger (≤ 64, P1: RR = 1.19, 95% CI (1.08, 1.32); P99: RR = 1.00, 95% CI (0.89, 1.12)) and women (P1: RR = 1.17, 95%CI (1.06, 1.30)). The effects of extremely low temperature on all types of emergency incidences were stronger than those of extremely high temperature in the whole year. In addition, impacts of cold weather lasted about several days while those of hot weather were acute and rapid. An increased frequency of emergency incidences is predicted by rising temperatures variations. These results have clinical and public health implications for the management of emergency incidences.