The loads regression estimator (LRE) was introduced by Wang et al. (2011) as an improved approach for quantifying the export of loads and the corresponding uncertainty from river systems, where data are limited. We extend this methodology and show how LRE can be used to analyze a 24 year record of total suspended sediment concentrations for the Burdekin River. For large catchments with highly variable discharge such as that of the Burdekin River, it is important to quantify loads and their uncertainties accurately to determine the current load and to monitor the effect of changes in catchment management. The extended methodology incorporates (1) multiple discounted flow terms to represent the effect of flow history on concentration, (2) a term that captures sediment trapping and spatial sources of flow in terms of the ratio of flow from above the Burdekin Falls Dam, and (3) catchment vegetation cover. Furthermore, we validated model structure and performance in relation to the application tested. We also considered errors in gauged flow rates of 10% that were consistent with the literature. The results for the Burdekin site indicate substantial variability in loads across years. The inclusion of vegetation cover as a predictor had a significant impact on total suspended sediment (TSS) concentration, with values up to 2.1% lower noted per increasing percentage of vegetation cover. TSS concentration was up to 38% lower in years with greater proportions of flow from above the dam. The extended LRE methodology resulted in improved model performance. The results suggest that management of vegetation cover in dry years can reduce TSS loads from the Burdekin catchment, and this is the focus of future work.