Sensitivity and uncertainty analysis of contaminant fate and transport modeling have received considerable attention in the literature. In this study, our objective is to elucidate the uncertainty pertaining to micropollutant modeling in the sediment-water column interface. Our sensitivity analysis suggests that not only partitioning coefficients of metals but also critical stress values for cohesive sediment affect greatly the predictions of suspended sediment and metal concentrations. Bayesian Monte Carlo is used to quantify the propagation of parameter uncertainty through the model and obtain the posterior parameter probabilities. The delineation of periods related to different river flow regimes allowed optimizing the characterization of cohesive sediment parameters and effectively reducing the overall model uncertainty. We conclude by offering prescriptive guidelines about how Bayesian inference techniques can be integrated with contaminant modeling and improve the methodological foundation of uncertainty analysis.