Extreme temperature events have affected Chinese city residents more frequently and intensively since the early 2000s,but few studies have identified the impacts of extreme temperature on mortality in different city clusters of China.This study first used a distributed lag,nonlinear model to estimate the county/district-specific effects of extreme temperature on nonaccidental and cardiovascular mortality.The authors then applied a multivariate meta-analysis to pool the estimated effects in order to derive regional temperature-mortality relationship in three large city clusters-the Beijing-Tianjin-Hebei (BTH) region,the Yangtze River Delta (YRD),and the Pearl River Delta (PRD),which represent north and south regions.With 0-3 days’ lag,the strongest heat-related mortality effect was observed in the BTH region (with relative risk (RR) of 1.29;95％ confidence interval (CI):1.13-1.47),followed by the YRD (RR =1.25;95％ CI:1.13-1.35) and the PRD (RR =1.14;95％ CI:1.01-1.28) areas.With 0-21 days’ lag,the cold effect was pronounced in all city clusters,with the highest extreme cold-related mortality risk found in the PRD area (RR =2.27;95％ CI:1.63-3.16),followed by the YRD area (RR =1.85;95％ CI:1.56-2.20) and BTH region (RR =1.33;95％ CI:0.96-1.83).People in the south regions tended to be more vulnerable to cold stress,but the north population was more sensitive to heat stress.By examining the effects of extreme temperature in city clusters of different regions,our findings underline the role of adaptation towards heat and cold,which has important implications for public health policy making and practice.