The Lewis diagram of a water molecule, at the bottom left, still holds, but MO theory identifies the two oxygen lone pairs explicitly as being in the 2s and 2p. orbitals. However, something is badly wrong with the simple localized MO theory as it has been presented here, because it predicts that the H-O- H bond angle will be 90 from the orientation of the p and p orbitals. This is reasonable for HS, for which the observed value is 92, but is wide of the mark for H0 with its measured 105 angle. One could give up the localized-bond approximation and go back to a full-molecule treatment of H0, but it is easier to introduce an idea that will be essential for carbon compounds: hybrid atomic orbitals.