We have initiated a study of improved methods for implementing dry (AIROX) processing of commercial spent nuclear fuels for recycling back into light-water reactors. In this proliferation-resistant recycle, the spent fuel is converted to a powder, blended with fresh medium-enriched uranium powder, and refabricated into fuel elements. Evaluations of neutronic characteristics show that it will be necessary to remove a substantial portion of the neutronabsorbing fission products in the spent fuel, especially lanthanides and rhodium, in order to achieve efficient utilization of the spent fuel. We have already modeled oxidative vaporization of selected fission products from the powder at 1000 °C. In addition to permanent gases and fission products that are vaporized during pellet sintering, this can remove Tc, Mo, and some Ru and improve the neutronics. A number of approaches are being evaluated for removing lanthanides and rhodium, initially by thermodynamic modeling and review of literature. The lanthanides exist in solid solution with UO2 so separations methods will require conversion to fine powder; rhodium is present as a metallic inclusion in the epsilon phase. Chlorination of finely powdered oxide at 1100-1200 °C would vaporize substantial portions of Nd, Eu, Gd, and Rh. A fraction of the uranium would also vaporize; if significant, it could be recovered for recycle. Magnetic and electrostatic methods were evaluated for separation of lanthanide from spent fuel. They are not likely to be practical. However, static separation techniques may be applicable for removing rhodium in the fine powder metallic inclusions. These technicalconsiderations provide the basis for a suggested experimental program.