Diffraction propagation over hills and ridges at VHF and UHF is commonly estimated using Fresnel knife edge diffraction. This approach has the advantage of simplicity, and for many geometries yields accurate results. However, since it neglects the shape and composition of the diffracting surface, it can in some cases yield results which are in serious disagreement with measurements. To remedy this, attempts have been made to approximate the diffracting hill or ridge by other shapes, most notably cylinders. These approaches have not been widely adopted, due in large part to their greater numerical complexity. In this paper it is proposed to apply wedge diffraction in the format of the geometrical theory of diffraction (GTD), modified to include finite conductivity and local surface roughness effects. It is shown that, for geometries with grazing incidence and/or diffraction angles, significant improvement in accuracy is obtained. Further, the GTD wedge diffraction form used is based on the Fresnel integral, so that it is only slightly more complex numerically than knife edge diffraction. Finally, the GTD includes reflections from the sides of the ridge (wedge faces), and can be extended to multiple ridge diffraction and three-dimensional terrain variations.