In northern Mississippi, roots in riparian zones were studied in an attempt to quantify the effects of root reinforcement of the soil matrix. The roots of trees can be treated as elastic-reinforced elements, and a function of the tensile strength of the roots can be added directly to the Mohr-Coulomb equation for failure criteria. Estimating root reinforcement and root-soil matrix interactions allows for the determination of whether bank vegetation is beneficial or detrimental. The research was conducted at the Goodwin Creek Experimental Watershed, located near Batesville, MS. This investigation quantifies root tensile strength of the sweet gum (Liquidamar syraciflua) in a cohesive, fine-grained, primarily loess-derived fluvial material. During the field research, trenches were excavated to gain access to the roots being studied. These trenches allowed mapping of the roots, as well as direct tensile testing of the roots. Increased tensile strength due to root reinforcement was found to be between 0.0 and 245 kPa, depending on depth. For a given depth of 40 cm, the increased tension due to root reinforcement averaged 148 kPa, depending on lateral distance from tree. A modified root reinforcement model was developed to explain the root-soil interaction observed at the research site. Itasca's Fast Lagrangian Analysis of Continua model was employed in determining the role of root reinforcement. The modeling results showed a contrast between root-reinforced and unreinforced soil. When no root reinforcement existed, the slope failed marginally. When simulated root reinforcement of 20 kPa was applied, the slope was shown to be completely stable.