In previous papers, a theory was proposed to explain the thermodynamic functions of substoichiometric fluorite structure compounds, on the basis of a new defect, which was called the tetrahedral defect, constituted by an oxygen vacancy which is bound with two reduced cations in its coordination tetrahedron. In this paper, the relevant energy parameters are discussed, and their number reduced to a “bonding energy” δ of the tetrahedral defect, a “strain energy” Δ measuring in first approximation, an isotopic strain introduced by packing tetrahedral defects in the fluorite lattice, a long-range electrostatic interaction u between “free” charged point defects and a “dipolar interaction energy” ϵ between tetrahedral defects: the latter two parameters can be in principle evaluated. By varying suitably the energy parameters, and comparing the results with the phase diagrams of oxides of the lanthanide and actinide elements, the difference between PuO 2−x (and AmO 2−x) (for which no low-temperature subphase is known) and all other fluorite-structure oxide systems may be explained. Variations in the slope of experimental oxygen potential vs stoichiometry curves which are found at high temperature for all these system, are explained (and fairly reproduced) in terms of residual structures.