Abstract Research on the critical shear stresses for erosion and deposition for cohesive sediment has attracted substantial attention from both engineering and theoretical viewpoints due to their importance in sediment transport theory. Previous studies have proposed a large number of empirical and semi-empirical methods to estimate the critical erosion and deposition shear stress, but comparative analyses and validation of the existing methods are still lacking, leaving questions regarding the applicability ranges of the methods. The current paper evaluates the performance and applicability range of five critical erosion shear stress methods derived from different hypotheses on sediment transport for flume experiments and natural tidal rivers using a process-based model. In addition, the effect of the critical deposition shear stress on sediment transport is investigated. The results show that the different critical erosion shear stress methods yield distinctly different prediction results, and their performance and applicability ranges are discussed by comparing their predictions with measured sediment concentrations from the Shenzhen River and measured geometric changes from the Partheniades' flume experiment. The hiding and exposure effect has been recognized as a crucial factor in the incipient motion of sediment on nonuniform beds. A sensitivity analysis of selective deposition and continuous deposition justifies the existence of the critical deposition shear stress. The current study highlights the performance and applicability ranges of the existing critical shear stress methods in sediment transport modeling for uniform and nonuniform beds, which will enrich understanding of the underlying mechanisms of erosion and deposition of cohesive sediment.