Most research to characterize the wind-erosion susceptibility and the degree of surface roughness required to suppress erosion on erodible surfaces has been empirical. However, a recently proposed shear velocity ratio model attempts to place shear stress partitioning in an entirely theoretical framework. The purpose of this study was to directly measure components of shear stress in a sparsely vegetated environment in order to evaluate the model. For the field study, new instrumentation was developed to measure drag on a creosote shrub, and Irwin sensors were modified to measure surface shear stress in the field. Simultaneous measurements of total shear stress and surface shear stress were taken at four sites of different roughness densities, in the Eldorado Valley, Nevada. Results indicate that porous shrubs had greater drag coefficients (Cd = 0.485) than did solid elements (sphere Cd = 0.3) and are more effective at protecting a surface. Values of β, the ratio of element to surface drag coefficients, were therefore higher than previously published values. Surface and total shear stress scaled consistently with each other at a range of wind speeds, and varied according to the roughness density of the surface. Shear stress partitioning values agreed well with previously published field data and some wind-tunnel data. The theoretical model predicted the results successfully when m = 0.16, where m is an empirical constant that accounts for the difference in average stress and the maximum surface stress in initiating erosion. The wide applicability of the model is likely due to the inclusion of the adjustable m, which accommodates all values of β and σ (ratio of roughness element basal area to frontal area).