Refractive spreading and attenuation limit the coverage of sub-bottom sonar by reducing both its angular resolution and amplitude, especially at angles near critical. As a means of overcoming these limitations, there has been great interest in post-critical penetration into sediments. One explanation for the reported post-critical penetration is rough interface scattering. In order to evaluate the importance of this effect, a simplified laboratory environment has been constructed to isolate rough interface scattering in an environment that excludes conditions necessary for alternate hypotheses. Accurate characterization of the medium is ensured by replacing the sediment with a homogeneous viscoelastic solid. The transducers are located in the water column, where the response can be well characterized. An automated positioning system and an interface area that is much larger then the correlation length of the roughness are employed to allow many independent measurements and thus, well determined statistics. A perturbation theory model has been implemented for comparison with experimental results and to assist in the selection of interface roughness spectra.