Computation of river suspended sediment loads is highly significant now-a-days for several reasons, such as predicting reservoir sedimentation, studying land erosion rates and understanding sediment transport to oceans. For achieving this, sediment rating curves which is a plot of suspended sediment concentration (C) vs. river discharge (Q), can be used instead of sampling programs which are sometimes insufficient to explain the continuous record..Many studies showing dynamics of suspended sediment have been introduced in the past literature, but mostly they have focused on C-Q analysis by plotting together all the data points from a catchment. This generally leads to poor correlation as C-Q relationship is often non-unique due to effects like seasonality and mixing of surface with subsurface water. In this study we therefore perform event-wise C-Q analysis, particularly considering recession flow curves and more emphasis has been given to three important recession characteristics: (1) power law exponent b, (2) dynamic C-Q relationship (explained by coefficient a) and (3) recession length (duration of a recession event). Long term data from fifty US basins that include at least ten numbers of recession events for each basin, are considered for the analysis. The results reveal that C-Q relationships for individual recession events are much stronger compared to those observed when all the data points are plotted together for a basin. We thus provide a new framework for characterizing C-Q relationships, which in turn will help us in addressing several practical issues related to soil loss and sediment movement and also to develop a Rating Curve with minimum error.