We report a comprehensive thermodynamic analysis of thermal oxidation-reduction cycles for producing hydrogen that use metal ferrites with the spinel structure (MFe2O4; M = Fe, Co, Ni, and Zn) as the redox material. Solution phases (both solid and liquid) were included in the calculations as well as the expected line compounds. Omitting solution phases, whose existence is experimentally well-documented, has a very significant impact upon the results of the calculations. Thermodynamic modeling of the three important material-related aspects of the process was performed, including synthesis of the ferrite materials from bulk oxides, thermal reduction at high temperatures, and reoxidation by reaction with steam. An experimental investigation of the NixFe3-xO4 system was performed to provide compositional data for comparison to model predictions. The results indicate that the Fe[Ni ratio, thermal reduction reaction kinetics, and the specifics of the cooling process affect the composition of the synthesized material. In particular, the presence of oxygen in the atmosphere during the cooling period following calcination substantially alters the sample composition. Predicted compositions following thermal reduction indicate that the stabilities are Fe3O4 > CoFe2O4 similar to NiFe2O4 > ZnFe2O4 and that the zinc-substituted ferrite is less desirable for solar hydrogen generation because of the high vapor pressure of zinc. Finally, modeling of the water oxidation step shows that efficient reoxidation to the original ferrite is thermodynamically feasible in all cases. We conclude that the temperature history and level of background O-2 present will affect both the phase purity of the initially formed material and the stability of the composition over the course of thermal cycling.