In the context of band gap engineering of the ZnO photoactive material for solar harvesting applications via W doping, a number of a priori discrepant experimental observations in the literature concerning ZnO c axis expansion/contraction, band gap red-or blue-shifting, the Zn-substitutional or interstitial nature of W atoms, or the W 6+ charge compensation view with ZnO native defects are addressed by thorough density functional theory calculations on a series of bulk supercell models encompassing a large range of W contents. The present results reconcile the observations, which are dissimilar at first sight, by showing that interstitial W (W i ) is only energetically preferred over substitutional (W Zn ) at large W doping concentrations; the W Zn c lattice expansion can be compensated by a W-triggered Zn-vacancy (V Zn ) c lattice contraction. The presence of W Zn energetically fosters nearby V Zn defects, and vice versa, up to a double V Zn situation. The quantity of mono-or divacancies explains the lattice contraction/ expansion, and both limiting situations imply gap states which reduce the band gaps, but increase the energy gaps. On the basis of the present results, the ZnO band gap redshifting necessary for solar light-triggered processes is achievable by W doping in Zn-rich conditions.