Abstract Based on monthly data from 1960 to 2007 of 740 stations in China, we investigate the response of monthly extreme high-temperature (MEHT) to atmospheric oscillation through spatial Generalized Pareto Distribution (GPD) model. Here we take different groupings of North Atlantic Oscillation (NAO) and Southern Oscillation (SO) as forcing parameters in the model and evaluate the response of MEHT to atmospheric oscillation in 9 different scenarios. Results show that the impact of NAO/SO on MEHT of China is significant and different responsen to different regions, i.e. the regions showing significant dependence of MEHT on atmospheric oscillation are Northeast China, the eastern part of the Tibetan Plateau, reaches of Yangtze and Yellow river and part of South China. To be specific, 1) MEHT obviously dependent on NAO is the north-east areas, followed by the eastern Tibet Plateau; MEHT obviously dependent on SO is mainly the Eastern Tibet Plateau and the Yangtze River and Yellow River and etc.; 2) Nonlinear addition effect of NAO and SO is not obvious, their impact on MEHT is mainly one-way, that is, both acting as forcing factors， most of the region forced by them is not significantly weakened or strengthened, and the reason may be related to differences in the atmosphere oscillation spatial and temporal distribution.