In contradiction to current bonding paradigms, two terminal Au-oxo molecular complexes have been synthesized by reaction of AuCl3 with metal oxide-cluster ligands that model redox-active metal oxide surfaces. Use of K10[alpha2-P2W17O61].20H2O and K2WO4 (forming the [A-PW9O34]9- ligand in situ) produces K15H2[Au(O)(OH2)P2W18O68].25H2O (1); use of K10[P2W20O70(OH2)2].22H2O (3) produces K7H2[Au(O)(OH2)P2W20O70(OH2)2].27H2O (2). Complex 1 crystallizes in orthorhombic Fddd, with a=28.594(4) A, b=31.866(4) A, c=38.241(5) A, V=34844(7) A3, Z=16 (final R=0.0540), and complex 2 crystallizes in hexagonal P6(3)/mmc, with a=16.1730(9) A, b=16.1730(9) A, c=19.7659(15) A, V=4477.4(5) A3, Z=2 (final R=0.0634). The polyanion unit in 1 is disorder-free. Very short (approximately 1.76 A) Au-oxo distances are established by both X-ray and 30 K neutron diffraction studies, and the latter confirms oxo and trans aqua (H2O) ligands on Au. Seven findings clarify that Au and not W is present in the Au-oxo position in 1 and 2. Five lines of evidence are consistent with the presence of d8 Au(III) centers that are stabilized by the flanking polytungstate ligands in both 1 and 2: redox titrations, electrochemical measurements, 17 K optical spectra, Au L2 edge X-ray absorption spectroscopy, and Au-oxo bond distances. Variable-temperature magnetic susceptibility data for crystalline 1 and 2 establish that both solids are diamagnetic, and 31P and 17O NMR spectroscopy confirm that both remain diamagnetic in solution. Both complexes have been further characterized by FT-IR, thermogravimetric analysis (TGA), differential scanning calorimetry (DSC), and other techniques.