Rutile TiO 2 (110) single crystals have been doped by nitrogen-ion implantation. The change in the valence band and in the core level peak shapes are characterized by photoemission spectroscopy. Surface morphologies are characterized by scanning tunneling microscopy. N-dopants are observed to be in a 3− charge state and to substitute for O-anions in the TiO 2 lattice for N-concentrations up to ∼5% of the anions. The higher valency of the N-dopants compared to the host O-anions is proposed to be compensated by the formation of O-vacancies and/or Ti-interstitials. Two chemically shifted components arise in the Ti-2p core level upon N-doping. These components, shifted by 0.9eV and 2.1eV, are assigned to Ti-bound to N-ligands and possibly due to O-vacancies in the lattice. The Ti-3d band gap state observed in UPS is initially suppressed upon room temperature N-implantation and recovers a similar intensity as for undoped TiO 2 samples upon annealing. This indicates that electrons left behind upon creation of O-vacancies are filling the N-2p level rather than Ti-3d states. The filled N-2p state is found at the top of the TiO 2 valence band and is believed to be responsible for the band gap narrowing of N-doped TiO 2 that shifts the photoactivity of TiO 2 into the visible spectrum.