Abstract: Based on daily minimum temperatures simulated by the COSMO-CLM (CCLM) for 1960-2100 and the landuse data in 2000 in China, the Intensity-Area-Duration (IAD) method was applied to characterize the extreme low-temperature events, the relationship between the intensity and spatial coverage of extreme low-temperature events, the spatial distribution of the most severe low-temperature events, and the exposure of farmlands to extreme low-temperature events under the global warming of 1.5℃ (RCP 2.6 scenario) and 2.0℃ (RCP 4.5 scenario). The results are as follows. (1) During the period of 1.5℃ warming, the frequency of extreme low-temperature events that can last for one to nine days will decrease by 30%-54%, the intensity will change by -1%-8.8%, and the area influenced by the extreme low-temperature events will decrease by 7%-21% compared to that during the reference period (1986-2005). For the 2.0℃ warming period, the frequency of the extreme low-temperature events will decrease by 48%-80%, the intensity will increase by 6%-11.5%, and the area affected by extreme low-temperature events will change by -14%-19%. (2) There are possibilities that both the intensity of and the area affected by the extreme low-temperature events in the future will exceed that of the most severe event in the reference period. Intensity of the most severe event in the context of 1.5℃ warming will be stronger than that in the context of 2.0℃ warming over the same area, but the center of the most severe event might move from Northwest and Southwest China to Central and South China. (3) Farmland exposure to the extreme low-temperature events will decrease in the warming periods than in the reference period, and the higher the temperature increase, the greater the decrease. However, farmland exposure to the most severe events will increase to a certain extent in East, North and Central China in the context of 1.5℃ warming, and such increases are even more obvious in East and North China in the context of 2.0℃ warming. Aforementioned findings indicate that frequency and coverage of the extreme low-temperature events will decrease, but their intensity might increase with the rising of temperature. With the increasing occurrence probability of the most severe event with higher intensity over larger area than that in the past, improvement of early warning is still imperative for mitigation of the societal and economic impact of extreme events.