Wind noise is a problem in seismic surveys and can mask the seismic signals at low frequency. This research investigates ground motions caused by wind pressure and shear stress perturbations on the ground surface. A prediction of the ground displacement spectra using the measured ground properties and predicted pressure and shear stress at the ground surface is developed. Field measurements are conducted at a site having a flat terrain and low ambient seismic noise. Triaxial geophones are deployed at different depths to study the wind-induced ground vibrations as a function of depth and wind velocity. Comparison of the predicted to the measured wind-induced ground displacement spectra shows good agreement for the vertical component but significant underprediction for the horizontal components. To validate the theoretical model, a test experiment is designed to exert controlled normal pressure and shear stress on the ground using a vertical and a horizontal mass-spring apparatus. This experiment verifies the linear elastic rheology and the quasi-static displacements assumptions of the model. The results indicate that the existing surface shear stress models significantly underestimate the wind shear stress at the ground surface and the amplitude of the fluctuation shear stress must be of the same order of magnitude as the normal pressure. Measurement results show that mounting the geophones flush with the ground provides a significant reduction in wind noise on all three components of the geophone. Further reduction in wind noise with depth of burial is small for depths up to 40 cm.