BACKGROUND: Non-optimal ambient temperature has detrimental impacts on mortality worldwide, but little is known about the difference in population vulnerability to non-optimal temperature and temperature-related morbidity burden between developing and developed countries. OBJECTIVES: We estimated and compared the associations of emergency department visits (EDV) with non-optimal temperature in terms of risk trigger temperature, the average slope of exposure-risk function and attributable risk in 12 cities from China and Australia. METHODS: We modelled the associations of EDV with heat during warm season and with cold during cold season, separately, using generalized additive model. Population vulnerability within a given region was quantified with multiple risk trigger points including minimum risk temperature, increasing risk temperature and excessive risk temperature, and average coefficient of exposure-risk function. Fraction of EDV attributable to heat and cold was also calculated. RESULTS: We found large between- and within-country contrasts in the identified multiple risk trigger temperatures, with higher heat and cold trigger points, except excessive risk temperature, observed in Australia than in China. Heat was associated with a relative risk (RR) of 1.009 [95% confidence interval (CI):1.007, 1.011] in China, which accounted for 5.9% of EDV. Higher RR of heat was observed in Australia (1.014, 95% CI: 1.010, 1.018), responsible for 4.0% of EDV. For cold effects, the RR was similar between two countries, but the attributable fraction was higher in China (9.6%) than in Australia (1.5%). CONCLUSIONS: Exposure to heat and cold had adverse but divergent impacts on EDV in China and Australia. Further mitigation policy efforts incorporating region-specific population vulnerability to temperature impacts are necessary in both countries.