An effective way to reduce the large quantities of Pu currently accumulated worldwide would be to use uranium-free fuel in light water reactors (LWRs) so that no new Pu is produced. Such a possibility could be provided by an LWR fuel consisting of Pu in a neutronically inert matrix. It may be necessary to add a burnable absorber or thorium to reduce the reactivity swing during burnup. The methods and data currently used for LWR analyses have not been tested in conjunction with such exotic fuel materials. An international exercise has accordingly been launched to compare the relative performance of different code systems and the accuracy of the basic data. Comparison of the results of cell calculations done with fixed isotopic densities against reference Monte Carlo results shows fairly small but systematic differences in the multiplication factors. A sensitivity analysis done with different basic cross section libraries and the same code system allows one to distinguish between the effects of the codes and those of the databases. The results of the burnup calculations indicate a fair agreement in kINF both at beginning of life (BOL) and after 1200 days of irradiation [end of life (EOL)] under conditions representative of a present-day pressurized water reactor. At BOL, the fuel temperature coefficients agree fairly well among the different contributions, but unacceptably large differences are observed at EOL. The void coefficients agree well for low voidage, but for void fractions 90%, there are significant effects mostly due to the databases used. The agreement in the calculated boron worths is good.