The interaction of NO with SrTiO3(100) surfaces was investigated using thermal desorption, photoemission, and first-principles density-functional calculations. The crystals used in the experiments exposed mainly (>80%) the TiO2-terminated face of SrTiO3(100). On the stoichiometric surfaces, the adsorption of NO was completely reversible at submonolayer coverages. Clear peaks for desorption of NO were found at 125 (multilayer state), 160, and 260 K, plus a long tail between 300 and 450 K. Desorption of N2O was detected only near 125 K with the multilayer of NO. DF calculations give adsorption energies of 14 and 6 kcal/mol for NO on the TiO2- and SrO-terminated faces of SrTiO3(100), which are consistent with the peaks at 260 and 160 K seen in thermal desorption. On the TiO2-terminated face of SrTiO3(100), there is substantial hybridization between the orbitals of NO and the oxide bands. This is not seen on the SrO-terminated face, where the large positive charge on the Sr sites leads to weak adsorption bonds. A reaction channel for the production of N2O and N-2 is opened by partially reducing the SrTiO3(100) surface. The cleavage of N-O bonds produces adatoms that quench vacancy states in the valence region and reduce the signals for Ti3+ and Ti2+ cations in core-level photoemission. DF calculations indicate that the adsorption of a NO single molecule over a vacancy site is a highly exothermic process (greater than or equal to70 kcal/mol) that leads to a large elongation (similar to0.20 Angstrom) but not a complete rupture of the N-O bond. The dissociation of this bond is facilitated by the addition of a second NO molecule and formation of an ON-NO dimer. The behavior of SrTiO3 illustrates the important effects that metal<---->oxygen<---->metal interactions can have on the electronic and chemical properties of a mixed-metal oxide. When dealing with the design or performance of ABO(3) perovskite catalysts, a simple extrapolation of the catalytic properties of the individual AO and BO2 oxides may not be a reliable approach. (C) 2003 American Institute of Physics.