The Geostationary Synthetic Aperture Radar (GEO SAR) concept aims at defining an Earth Observation system able to provide regional coverage with large swaths and subcontinental access with a very short revisit time by exploiting the unique characteristics of the GEO orbit. These peculiar characteristics make GEO SAR suitable for imaging and interferometry of fastevolving large-scale phenomena, such as ground motion or water  vapor. The design of the orbit for a GEO SAR is discussed to provide, at one time, wide imaging capability, compensation for the huge power loss due to the distance, and short interferometric revisit. In particular, interferometry imposes constraints on the repeatability of the orbit that are far more restrictive than the control required for telecommunication satellites. Here, two different specific control approaches are proposed to ensure small normal baselines and to maximize angular band overlap between subsequent acquisitions. The first one is based on four control maneuvers per orbit period that exploits tangential, normal, and radial burns, whereas the second one reduces to three the required daily maneuvers that are performed only in tangential and normal directions. The latter assures accurate motion repetition while saving the required ΔV .