Forecasting River discharge is always a serious problem waiting to be solved. The flooding in Ganges-Brahmaputra downstream has caused numerous lost of lives and assets in recent years. The influence of tropical ocean SST variation like ENSO to global climate has been on focus in recent two decades. But the study of Bangladesh flooding relate to tropical SST is just beginning. The rainfall in Ganges valley is mainly caused by summer monsoon which is driven by sea-land temperature difference. Correlation analysis based on 50 years (1950-1999) shows that Pacific Ocean SST has significant linear correlation with discharge of Ganges than Indian Ocean. The western and mid part of the equatorial Pacific Ocean take the highest coefficients from simultaneous to two months lead time. At the years when La Nina (El Nino) is developing in winter, the preceding summer discharge of Ganges is significant high (low). Scattered points map and lag correlation suggests that May, June, July SST in Mid-Pacific Ocean (Coral Sea) is significant negative (positive) linear related with July and August discharge of Ganges while Western Pacific SST only shows the simultaneous positive relationship. Flood forecasting of two months lead by SST is possible. The discharge of Brahmaputra River shows no linear relation to equatorial SST variation. The internal mechanisms behide the correlations need to be unveiled.