Previous authors have argued that the alloying additions that protect zirconium from nodular corrosion can do so because they have stable oxidation states in zirconia other than M 4+. However, the arguments used to predict the oxidation states of such alloys are incorrect. This paper shows how the oxidation states of dopants in an oxide should be calculated and calculates the true oxidation states of the various alloying additions in zirconia. The resulting oxidation states are often quite different from those suggested previously. The assertion that protective alloys produce aliovalent oxidation states is nevertheless vindicated. We discuss the implications for these results on the suggested mechanisms of oxide growth and nodular corrosion.