Large regions of East Antarctica lack a reasonable topographic model because, until recently, only a few observations of ice thickness have been available to constrain the bedrock elevation. The acquisition of GRACE satellite gravity data has created a new opportunity to model the sub-ice topography. Here we have applied two methods for predicting topography based on the satellite data. Gravity inversion is a classical geophysical technique that predicts topography based on the physics relating it to gravity. Cokriging is a statistical method that uses the spatial covariance between datasets to predict one in the absence of the other. The geophysical and statistical solutions are compared to the best-known topography model (BEDMAP) in an area that is relatively well constrained by the BEDMAP data coverage.