This paper examines the credibility and predictability of sediment flux of the Changjiang River that has discharged into the seas on the basis of historical database. The assumption of the study stands on the lack of sufficient observation data of suspended sediment concentration (SSC) during peaking flood period, which most likely results in the application of an inappropriate method to the downstream-most Datong hydrological gauging station in the Changjiang basin. This insufficient method (only 30-50 times of SSC observation per year), that obviously did not cover the peaking SSC during peaking floods, would lead to an inaccuracy in estimating the Changjiang sediment load by 4.7×10⁸ t/a (multiyearly) into the seas. Also, sediment depletion that often takes place upstream of the Changjiang basin has, to some extent, lowered the credibility of traditional sediment rating curve that has been used for estimating sediment budget. A newly-established sediment rating curve of the present study is proposed to simulate the sediment flux/load into the seas by using those SSC only under discharge of 60000 m³/s at the Datong station-the threshold to significantly correlate to SSC. Since discharge of 60000-80000 m³/s is often linked to extreme flood events and associated sediment depletion in the basin, un-incorporating SSC of 60000-80000 m³/s into the sediment rating curve will increase the credibility for sediment load estimation. Using this approach of the present study would indicate the sediment load of 3.3×10⁸-6.6×10⁸ t/a to the seas in the past decades. Also, our analytical result shows a lower sediment flux pattern in the 1950 s, but higher pattern in the 1960 s-1980 s, reflecting the changes in land-use in the upstream of Changjiang basin, including widely devastated deforestation during the middle 20th century.