Glacial isostatic adjustment (GIA) represents a source of uncertainty for ice sheet mass balance estimates from the Gravity Recovery and Climate Experiment (GRACE) time-variable gravity measurements. We evaluate Greenland GIA corrections from Simpson et al (2009 Quat. Sci. Rev. 28 1631–57), A et al (2013 Geophys. J. Int. 192 557–72) and Wu et al (2010 Nature Geosci. 3 642–6) by comparing the spatial patterns of GRACE-derived ice mass trends calculated using the three corrections with volume changes from ICESat (Ice, Cloud, and land Elevation Satellite) and OIB (Operation IceBridge) altimetry missions, and surface mass balance products from the Regional Atmospheric Climate Model (RACMO). During the period September 2003–August 2011, GRACE ice mass changes obtained using the Simpson et al (2009 Quat. Sci. Rev. 28 1631–57) and Aet al (2013 Geophys. J. Int. 192 557–72) GIA corrections yield similar spatial patterns and amplitudes, and are consistent with altimetry observations and surface mass balance data. The two GRACE estimates agree within 2% on average over the entire ice sheet, and better than 15% in four subdivisions of Greenland. The third GRACE estimate corrected using the (Wuet al 2010 Nature Geosci. 3 642–6)) GIA shows similar spatial patterns, but produces an average ice mass loss for the entire ice sheet that is 64 − 67 Gt yr−1 smaller. In the Northeast the recovered ice mass change is 46–49 Gt yr−1 (245–270%) more positive than that deduced from the other two corrections. By comparing the spatial and temporal variability of the GRACE estimates with trends of volume changes from altimetry and surface mass balance from RACMO, we show that the Wuet al (2010 Nature Geosci. 3 642–6) correction leads to a large mass increase in the Northeast that is inconsistent with independent observations.