The occurrence of summer high temperature is the comprehensive reflection of joint actions of several factors， in which atmospheric circulation background plays an important role in hightemperature events. According to the study of airsea interaction, there are maybe some feedback relations between SST change and atmospheric circulation, which determine the certain predictability of land temperature using SST. Based on this principle, with the 30year (1978—2007) West Pacific SST, and the numbers of high temperature days of different time, the strong signal sea area related to average high temperature days of different regions of Jiangsu is explored by using regional average method, field correlation analysis and optimization correlation processing technology. Setting West Pacific SST as the longterm forecast factor, the longterm forecast models of summer monthly high temperature days in different areas of Jiangsu are established. These models all pass through the significance level of α=0.01, and the effects of fitting are good. Experiments of practical forecast are carried out with good effects, and can be put into meteorological operations. The study has important guiding significance for service on the longterm forecast of high temperature and high temperature disasters in Jiangsu Province.