The DFT B3LYP/SBKJC method has been used to calculate the gas-phase optimized geometries of the glycolate oxoperoxo vanadium(V) complexes [V(2)O(2)(OO)(2)(gly)(2)](2-), [V(2)O(3)(OO)(gly)(2)](2-) and [VO(OO)(gly)(H(2)O)](-). The (51)V, (17)O, (13)C and (1)H chemical shifts have been calculated for the theoretical geometries in all-electron DFT calculations at the UDFT-IGLO-PW91 level and have been subsequently compared with the experimental chemical shifts in solution. In spite of being applied to the isolated molecules, the calculations allowed satisfactory reproduction of the multinuclear NMR solution chemical shifts of the complexes, suggesting that the theoretical structures are probably close to those in solution. The effects of structural changes on the (51)V and (17)O NMR chemical shifts have been analysed using the referred computational methodologies for one of the glycolate complexes and for several small molecules taken as models. These calculations showed that structural modifications far from the metal nucleus do not significantly affect the metal chemical shift. This finding explains why it is possible to establish reference scales that correlate the type of complex (type of metal centre associated with a certain type of ligand) with its typical region of metal chemical shifts. It has also been found that the V[double bond, length as m-dash]O bond length is the dominant geometrical parameter determining both delta(51)V and the oxo delta(17)O in this kind of complex.