We report on an investigation into the use of differential interferometric synthetic aperture radar (SAR) (DInSAR) for the discrimination between surface and subsurface features in a soil, undertaken at the Ground-Based SAR Microwave Measurement Facility. A temporal sequence of C-band VV SAR images of a drying soil containing a buried target was collected. While the phase record of the signal identified with the soil return showed almost no variation, in stark contrast, the phase from the buried target showed a strongly linear change with time. A model is presented, which describes the observed phase changes in terms of retardation of the signal by the soil dielectric properties, which are dependent upon the moisture content. The model confirms a strongly linear relationship between phase and volumetric soil moisture. The linearity promises to greatly simplify any exploitation scheme, and such a DInSAR scheme would be applicable at large standoff distances from airborne and spaceborne platforms, in contrast to current subsurface techniques which rely on close-in measurement to spatially isolate returns vertically in backscatter.