Rolling contact fatigue (RCF) is one of the major sources of subsurface-initiated spalling in bearing components. These types of spalls initiate in regions where microcracks are formed due to a localized increase in stress from inclusions near the surface. Ultrasonic bulk inspection is a powerful method to detect defects in a large volume of steel. However, inclusions near the surface of a sample can elude detection because reflections from the inclusion can be masked by reflections from the sample surface. This limitation can be eliminated if ultrasonic surface wave methods are used for inspection. Surface waves have material displacements localized near the sample surface but decay with depth giving an effective inspection of depth on the order of the wavelength. Ultrasonic scattering from inclusions also is wavelength dependent and these two aspects can complicate the interpretation of ultrasonic experimental data. In this presentation, a model is described for the scattering of a surface wave by a subsurface spherical inclusion. The amplitude-versus-depth profile of a surface wave is combined with the solution for the scattering of a shear wave from a spherical scatterer in order to approximate the problem of interest. Trends of reflected amplitude with respect to inspection frequency, inclusion depth, and inclusion diameter are discussed first. Then a necessary calibration experiment is described that uses subsurface defects of known size created using femtosecond laser machining. A model of the calibration sample allows measurements on unknown samples to be interpreted quantitatively. The final analysis shows that the reflected amplitude from multiple frequency measurements can be used to characterize the size and depth of the subsurface inclusions.