A fundamental precept of geochemistry is that arsenate coordinates at mineral surfaces in a predominately bridging-bidentate fashion. We show that this is incorrect for the model system, arsenate adsorbed at the surface of goethite (alpha-FeOOH), using a combination of XRD, EXAFS, and IR spectroscopic results. We report the crystal structure of pentaamminecobalt(III) arsenate, which consists of monodentate-coordinated metal-arsenato complexes that have Co-As distances of only 3.25 A. This result implies that metal-arsenic distances are not diagnostic for the coordination mode of arsenate. We show that the K-edge EXAFS spectra of pentaamminecobalt(III) arsenate and arsenate-goethite surface complexes are strikingly similar, which suggests that arsenate could be coordinated at the goethite surface in a monodentate fashion. Refinements of the k(3)-weighted EXAFS spectra of arsenate adsorbed on goethite results in values of CN(As-Fe) between 0.8-1.1 (+/-0.7), and there is no evidence that the coordination mode of arsenate changes as a function of pH or arsenate surface coverage. We report IR spectra from the first simultaneous IR and potentiometric titration of arsenate adsorbed on deuterated goethite (alpha-FeOOD) in D(2)O, and we show for the first time the As-O stretching bands of arsenate-goethite surface complexes. We deduce that arsenate-goethite surface complexes are un-, singly, or doubly protonated, depending on pH, from a principal component analysis of the As-O stretching region and an interpretation of the Type-B OH stretching region. In summary, our cumulative results show that arsenate coordinates at the water-goethite interface in a predominately monodentate fashion. Furthermore, we find no evidence for bridging-bidentate coordination, which is a finding that impacts oxoanion bioavailability and challenges theories of mineral dissolution and surface complexation.