Experiments conducted in the gas phase have led to the formation of a series of stable gold(II) complexes with nitrogen- and oxygen-containing ligands. Such complexes are very rare in condensed-phase chemistry. However, there is also a significant group of potential ligands, for example, H2O and NH3, for which stable complexes could not be formed. There an strong similarities between these observations and earlier results presented for silver(II), but both meter ions behave markedly different from copper(II). As a group the majority of successful gold(II) ligands are characterized by being good sigma donor-pi acceptor molecules; however, it is also possible to understand the ability of individual ligands to stabilize the metal ion in terms of a simple electrostatic model. Application of the latter reveals a semiquantitative trend between the physical properties of a ligand, e.g. ionization energy, dipole moment, and polarizability, and the ligand's ability to stabilize either Cu(LI), Ag(II), or Au(II). The model successfully accounts for the preference of Cu(LI) for aqueous chemistry, in comparison to the complete absence of such behavior on the part of Ag(II) and Au(LT). Ligands from recent examples of stable condensed-phase gold(II) complexes appear to meet at Least one of the criteria identified from the model.