Abstract. In the French context, it appears that the accuracy of the nuclear data comply with the utility need. However, a constant reduction of the calculation uncertainties is achieved with more precise physics and numerical methods and improved nuclear data. The aims for this uncertainty reduction are: to increase the competitiveness, by reduced operating conservatisms in order to extend the reactor lifespan and by optimising the existing series of nuclear plants and associated fuel cycle: extended burn-up, MOX fuel, reduced damage to the reactor vessel. . . and to evaluate and to develop the future nuclear energy systems (Generations III and IV) with a sufficient accuracy. The nuclear reactor operation is currently supported by numerical simulation: if the discrepancies between precalculated values and measures in operation are sufficiently low, the required authorisations are given by the Safety Authority. EDF core calculations have been relying on JEF2.2 for more than fifteen years. This association of code and database gives a satisfying uncertainty level for the current reactors and associated fuel management, but the development of a new calculation scheme named N3C-V2 has been decided at EDF, relying simultaneously on: -improved numerical scheme allowed by today's powerful computers and recent physics improvements; -a new set of nuclear data: JEFF3.1. This work is considerable, and the expected results are the reduction of uncertainty on the main parameters of future optimised fuel management in existing reactors and in EPR. Concerning Gen IV reactors and actinide transmutation, progresses are still needed (even for Sodium cooled FBR which are the reference Gen IV solution of the industrial EDF), mainly in the fast spectrum. The aim is to evaluate with a better accuracy: Fuel evolution, in particular, the formation of Minor Actinides (Am, Cm, Np) or important isotopes, like 232 U in the thorium cycle; Breeding gain ( 238 U and 232 Th captures, 239 Pu and 233 U fissions, in the fast range for 238 U and 239 Pu, in all ranges for 232 Th and 233 U); Doppler effect; Void effect in fast spectrum reactors is a major issue and can be influenced by the heavy nuclei cross sections, but also by the elastic and inelastic scattering cross sections of the coolant (scattering of Na or Pb for ex.). Finally, uncertainty on cross-sections must become available in the evaluation files, for they will help to differentiate the part of uncertainty coming from the data and from the calculation methods.