Abstract: In view of return intervals, we studied the long range correlation of extreme high temperature and low temperature in China and its relationship with group-occurring. It is found that the time series of extreme temperatures shows a long range correlation and the scaling exponent, which means the significance of long range correlation, is related to the atmospheric circulation. When the mainland China is under the control of a stable westerly circulation (e.g. in winter and summer), the temperature shows significant long range correlation, i.e. a greater scaling exponent, and the reconstructed events are usually group occurring. The groupoccurring of the return intervals of extreme events is defined in this paper according to this character. Through the study of extreme high and low temperatures, we found that when a time series shows a long-range correlation, its extreme temperature events show a group-occurring phenomenon. The scaling exponent of the original temperature has a better correspondence in spatial distribution to the groupoccuring index of extreme temperatures. And the long range correlation of time series may be the fundamental reason of groupoccurring of extreme events. From the perspective of inter annual change, the growth of groupoccurring index of extreme high temperature’s levelⅠregion tended to moderate, while that of the extreme low temperature’s tended downward. This is consistent with global warming in recent decades. And in the decadal time scales, the evolution characteristics of the groupoccurring index’s distribution is apparent. The overall area change was not obvious as to the extreme high temperature’s level Ⅰgroup occurring region, but the area of the extreme low temperature’s levelⅠ groupoccurring region reduced significantly. Therefore, the weakening of extreme low temperature events’ group occurring is very likely to contribute to the inter-decadal climate warming