Groundwater is a significant component of the hydrological cycle in arid and semi-arid areas. Its evapotranspiration is an important part of the water budget because many plants are groundwater-dependent. To restore the degraded ecosystems, the need is pressing to further our understanding of the groundwater evapotranspiration(ET_g) in arid and semi-arid areas. This study employed the White method to estimate ET_g at four sites in the Mu Us Sandy Land in northern China, and the four sites are covered by Salix psammophila(SP site), Artemisia ordosica(AO site), Poplar alba(PA site), and Carex enervis(CE site), respectively. The depth of groundwater table and the duration of drainage were taken into account in calculating the specific yield(S_y) to improve the accuracy of the ET_g estimats. Our results showed that from late May to early November 2013 the ET_g were 361.87(SP site), 372.53(AO site), 597.86(PA site) and 700.76 mm(CE site), respectively. The estimated ET_g rate was also species-dependent and the descending order of the ET_g rate for the four vegetation was: C. enervis, P. alba, A. ordosica, and S. psammophila. In addition, the depth of groundwater table has an obvious effect on the ET_g rate and the effect varied with the vegetation types. Furthermore, the evapotranspiration for the vegetation solely relying on the water supply from unsaturated layers above the groundwater table was much less than that for the vegetation heavily relying on the water supply from shallow aquifers. Its evapotranspiration is an important part of the water budget because an important part of the water budget because many plants are groundwater-dependent. To restore the degraded ecosystems, the need is pressing to further our understanding of the groundwater evapotranspiration (ET_g) in arid and semi-arid areas. This study employed the White method to estimate ET_g at four sites in the Mu Us Sandy Land in northern China, and the four sites are covered by Salix psammophila (SP site), Artemisia ordosica (AO site), Poplar alba (PA site), and Carex enervis (CE site), respectively. The depth of groundwater table and the duration of drainage were taken into account in the calculation of the specific yield (S_y) to improve the accuracy of the ET_g estimats. Our results showed that from late May to early November 2013 the ET_g were 361.87 (SP site), 372.53 (AO site), 597.86 (PA site) and 700.76 mm (CE site), respectively. The estimated ET_g rate was also speci es-dependent and the descending order of the ET_g rate for the four vegetation was: C. enervis, P. alba, A. ordosica, and S. psammophila. In addition, the depth of groundwater table has an obvious effect on the ET_g rate and the effect varied with the vegetation types. Furthermore, the evapotranspiration for the vegetation solely relying on the water supply from unsaturated layers above the groundwater table was much less than that for the vegetation heavily relying on the water supply from shallow aquifers.