Tele-connections between El Nino/Southern Oscillation (ENSO), North Atlantic Oscillation (NAO), Indian Ocean Dipole (IOD), Pacific Decadal Oscillation (PDO) and Atlantic multi-decadal oscillation (AMO) and flood-waterlogging disaster of different levels in agriculture of China were analyzed. Results show that: (1) the flood-waterlogging disaster is increasing with positive PDO and IOD in the last year on northeast and south of China, while decreases with positive ENSO in the last year on southwest of China. However, the flood-hazard crop area of southeast coastal areas of China was influenced significantly by a variety of climate indexes. (2) The intensities of correlation between the temporal pattern of the REOFs and climate indices with significant level tend to strengthen and to maintain a good stability in recent decades on different time scales, and thus is useful to predict agricultural flood disasters. ENSO has a good relationship with flood-waterlogging disaster and the stability of the relationship is best among all, and so it is the most suitable as potential climate index for the prediction of agricultural flood disasters. (3) Climate indices have the same direction (reinforcing each other) or opposite direction (weakening each other) of influences on flood-water logging disaster in different cool or warm events/periods.