Ocean tide models around Antarctica are presently only sparsely tested against independent data. Ocean tide modeling errors, along with subsequent ocean tide loading (OTL) displacement modeling errors, alias into altimetry and time variable gravity (e.g., Gravity Recovery and Climate Experiment (GRACE)) time series, for example. To validate various ocean tide models around Antarctica, GPS data from 15 sites have been used to derive three-dimensional displacement estimates at eight diurnal and semidiurnal tidal frequencies. Using hundreds of days of GPS data, harmonic parameters were estimated on a daily basis then combined. These were then compared with OTL displacement estimates derived from global and regional ocean tide models. In East Antarctica, where the tides are well defined, submillimeter differences are demonstrated in each coordinate component with the lunar N2 and Q1 constituents in closest agreement. As found in other studies, K1 and, especially, K2 agree less well. The spatial variation in the misfits for these two constituents indicates a site dependency, with the K2 errors also suggesting an interaction with satellite-dependent effects. In West Antarctica, where sites are nearer the largest ice shelves, agreement with the older models (CSR3 and TPXO.2) and NAO.99b is poor for all constituents. Modeled tidal gravity variations were also compared with gravity measurements at the South Pole. Overall the GPS and gravity data agree best with newer tide models, namely, TPXO.6, CADA00.10, FES99, and CATS02.01. However, validation data are lacking at the southern extents of the large ice shelves, and hence some uncertainty still exists in all ocean tide models in these regions.