Abstract: The flood and drought in Changjiang and Huaihe river (Jianghuai) areas are frequent in summer, especially in July and June, so forecasting the summer flood and drought in Jianghuai areas is always one of key points investigated by meteorologists in China. The previous studies more focused on the skin temperature anomalies in a single area, and paid less attention to the connections between the anomalies of sea-land thermal contrast in northern Africa areas and the flood and drought in Jianghuai areas in summer. In this paper, we investigate such relationships and obtain some preliminary results.By use of the NCEP/NCAR monthly mean re analysis data and the rainfall data at 743 stations in China, the Northern Africa areas are selected as the key regions according to the interdecadal variability characteristics of the flood and drought index (FDI) during 51 years(1954-2004) in Jianghuai areas in summer. Correlation analysis show that the surface temperature anomalies in key region have good continuity in winter, and the winter North Atlantic Oscillation (NAO) maybe is one of important reason which resulting the anomalies continuity. By singular value decomposition (SVD) analysis between the skin temperature in previous winter in Northern Africa areas and the summer rainfall in Jianghuai areas, it is found when the Northern Africa land is colder(warmer) and its northwestern sea is warmer(colder), the rainfall wholly increases (decreases) in the Jianghuai areas in summer. Further analysis finds that the anomaly of surface temperature contrast between sea and land has better indication than that in any single region in Northern Africa areas for forecasting the flood and drought in Jianghuai areas in summer, and therefore, a sea and land thermal contrast index (SLTCI) is defined to reflect the intensity of the large scale sea-land thermal contrast. Correlation analysis show the very positive correlation between the SLTCI and the FDI in summer in Jianghuai areas, and it can well indicate the extreme flood and drought years. So the index could be used to predict the wholly flood and drought in Jianghuai areas in summer.