A novel processor for the generation of a digital terrain model (DTM) using ground-based synthetic aperture radar (GB-SAR) imagery is presented. The performance of the processor has been assessed using the archive of C-band GB-SAR images available at the Joint Research Centre from the Sion Valley (CH) test site, from which a high-resolution ground-truth DTM is available. In order to reduce the standard deviation of the differences between the two DTMs, several techniques have been applied: interferogram averaging, phase referencing, and differential path calibration. In addition, different formulations have been reviewed, and its impact has been quantified in the DTM accuracy obtained. Results show that the standard deviation of the differences between the ground-truth DTM and that of the C-band GB-SAR can be reduced from the 5 m that was previously reported down to 3 m with the new processor.