The Gravity field and steady-state Ocean Circulation Explorer (GOCE) dedicated satellite gravity field mission was launched by the European Space Agency (ESA) on March 17, 2009. Since then, several Global Geopotential Models (GGMs) have been released based on data collected by GOCE. This paper evaluates different GOCE-only and GOCE/GRACE GGMs using 567 available GPS/Levelling points and terrestrial free-air gravity anomalies in Argentina. EGM2008 and EIGEN-51C models are also included in the evaluation procedure, even though they are not based on GOCE data. The idea of using these models is to identify any improvements of the new GOCE-only or GOCE/GRACE GGMs in Argentina. The truncation of the Stokes coefficients at a certain spherical harmonic degree n1 produces an omission error. In order to reduce this omission error, synthetic GGMs are evaluated by adding signal from EGM2008 and topographic effects through a Residual Terrain Model (RTM) to the satellite GGMs models. The evaluation is performed with an incremental step of one in harmonic degree, so that the most detailed possible evaluation of the GOCE-only and GOCE/GRACE GGMs will be performed. The results show that EGM2008 is better than all GGMs, used for evaluation in this study, in terms of the standard deviation of the geoid heights are concerned. This superiority is marginal and statistically insignificant, being at the 3-2 mm level. GOCE/GRACE GGMs are significantly better than EGM2008 in terms of the range of the differences with the GPS/Levelling data, since they reduce the 1.964 m of the EGM2008 range by as much as 0.21 m for DIR_R5. Contrary to the results for the validation with the GPS/Levelling data, all the GOCE and GOCE/GRACE GGMs are better at the 0.1 mGal level for the gravity anomaly differences.Facultad de Ciencias Astronómicas y Geofísicas