A hypothesis used to explain the relationship between timber harvesting and landslides is that tree roots add mechanical support to soil, thus increasing soil strength. Upon harvest, the tree roots decay which reduces soil strength and increases the risk of management -induced landslides. The technical literature does not adequately support this hypothesis. Soil strength values attributed to root reinforcement that are in the technical literature are such that forested sites can't fail and all high risk, harvested sites must fail. Both unstable forested sites and stable harvested sites exist, in abundance, in the real world thus, the literature does not adequately describe the real world. An analytical model was developed to calculate soil strength increase due to root reinforcement. Conceptually, the model is composed of a reinforcing element with high tensile strength, i.e. a conifer root, embedded in a material with little tensile strength, i.e. a soil. As the soil fails and deforms, the reinforcing element also deforms and stretches. The lateral deformation of the reinforcing element is treated analytically as a laterally loaded pile in a flexible foundation and the axial deformation is treated as an axially loaded pile. The governing differential equations are solved using finite-difference approximation techniques. The root reinforcement model was tested by comparing the final shape of steel and aluminum rods, parachute cord, wooden dowels, and pine roots in direct shear with predicted shapes from the output of the root reinforcement model. The comparisons were generally satisfactory, were best for parachute cord and wooden dowels, and were poorest for steel and aluminum rods. A parameter study was performed on the root reinforcement model which showed reinforced soil strength increased with increasing root diameter and soil depth. Output from the root reinforcement model showed a strain incompatibility between large and small diameter roots. The peak increase in soil strength attributed to roots was controlled by the small (<4mm) diameter root fraction. These results were used to calculate the effect of timber harvesting on a small, approximately 7.6 m³ (10 yd³), hypothetical landslide in a shallow, cohesionless, forest soil. The root reinforcement model predicted a post-harvest reduction in soil strength of 14 and 19 percent for a soil with and without 5 kPa (105 lbs/ft^2) of cohesion, respectively.