As part of the Cryosat Cal/Val activities and the International Partnerships in Ice Core Sciences (IPICS) pre-site survey ground based kinematic GPS measurements were conducted in early 2007 in the vicinity of the German overwintering station Neumayer. The investigated areas are the ice caps of Halvfarryggen and Søråsen, which rise from the Ekstrøm Ice Shelf to about 650 m surface elevation, and have an aeral extent of about 100 km x 50 km. Available digital elevation models (DEM) from radar altimetry and the Antarctic Digital Database show differences of several tenths of meters to each other in this coastal area, which necessitated accurate survey of the conditions on-site. Besides the main objectives, the surface topography is required for ice dynamic modelling and mass balance studies. The field program also included a ground-penetrating radar (GPR) survey and firn and ice core measurements in order to obtain height related spatial information on snow morphology. The GPS receiver was mounted on a snow vehicle which was navigated at a velocity between 10 and 12 kmh-1 along pre-defined tracks. The along track distance between data points is about 3 m. The GPS data were processed with nearby temporary reference stations, precise ephemeris and ionosphere-free solution. A GPS reference station was located at the seismological observatory Watzmann (70.93°S and 7.39°W) for the complete duration of GPS measurements. Local reference stations were positioned at four camps sides to shorten the baselines. Because of the relative short baselines, the derived elevation data show in general a high accuracy, which is checked with a crossover-point analysis. A complete DEM of the Halvfarryggen and Søråsen area is derived by a combination of GPS elevations, interferometric SAR (INSAR) and ICESat Laser data. GPS and ICESat data are used as ground control points in order to confine the precise satellite orbits for ERS. The INSAR derived DEM contains noise, depending on the coherence of the interferogram. An appropriate filter is applied in comparison to GPS height. The resulting INSAR height show a good overall agreement, but differences may depend on unsufficient differential correction, or varying snow pack properties and hence penetration depths of the radar signal. Finally, the resulting elevation model is used as a reference to explore the accuracy of several other DEMs in the area of investigation. For example, significant elevation differences between the here oresented DEM and the RAMP DEM can be seen at higher elevated parts. Differences due to the slope induced error of radar altimeters is evident in the steeper and lower elevated regions.