[en] When the residence time of nuclear fuel rods exceeds a given threshold value, several properties of the pellet material suffer changes and hence the posterior behaviour of the rod is significantly altered. Structural modifications start at the pellet periphery, which is usually referred to as rim zone. It is presently believed that these changes are a consequence of the localized absorption of epithermal neutrons by ²³⁸U, which effective cross section presents resonant peaks. Due to the chain of nuclear reactions that take place, several Pu isotopes are born especially at the rim. In particular, the fissile character of ²³⁹Pu and ²⁴¹Pu is the cause of the increased number of fission events that occur in the pellet periphery. For this reason, the power generation rate and the burnup adopt a non uniform distribution in the pellet, reaching at the rim values two or three times higher than the average [1]. The rim zone starts to form for a burnup threshold value of about 50-60 MWd/kgHM and its width increases as the irradiation progresses. The microstructure of this zone is characterized by the presence of small grains, with a typical size of 200 nm, and large pores, of some μm. Even though the rim zone is very thin, it has a significant effect on the mechanical integrity of the pellet, particularly when it makes contact with the cladding, and on the temperature distribution in the whole pellet, because of its low thermal conductivity [1,2]. The numerical codes designed to simulate fuel behaviour under irradiation must include the phenomena associated to high burnup if they aim at extending the prediction range, and this is the purpose with our DIONISIO code. But a detailed analysis of the phenomena that take place in this region demands the use of neutronic codes that solve the Boltzmann transport equations [3] in a number of energy intervals (groups), including adequate considerations in the region of the resonant absorption peaks of ²³⁸U. These cell codes predict with high precision the neutron flux, burnup and concentration of every isotope, fissile, fissionable or fertile, gaseous or solid, all of them as functions of radius and time. But this formidable task is not suitable to be included in a fuel performance code, which must attend the great number of thermomechanical and thermochemical processes within the fuel rod. To accommodate both requirements, a simplified treatment is adopted consisting of restricting the balance equations to more relevant nuclides and reducing the energy spectrum to a single group. The purpose is to obtain empirical expressions to represent, with the higher possible approximation degree, the absorption, capture and fission cross sections of these isotopes as functions of the initial enrichment in ²³⁵U, the average burnup and the radial coordinate. The curves obtained with a so drastic simplification demand a careful testing before incorporation in the general fuel behaviour code. This testing is performed via comparison with the reliable reactor codes. The first antecedent in this type of analysis is found in the RADAR model [4] which was validated against the WIMS [5,6] code. The TUBRNP model, included in the TRANSURANUS code [7] and the RAPID model [8] are also based on the same concept. In this work curves fitted for the cross sections of ²³⁵U, ²³⁶U, ²³⁸U, ²³⁹Pu, ²⁴⁰Pu, ²⁴¹Pu and ²⁴²Pu are obtained from the predictions of the reactor cell codes HUEMUL [9] and CONDOR [10] for an average burnup ranging from fresh fuel to 120 MWd/kgHM and for an initial enrichment ranging from natural uranium to 12%. The final purpose is to extend the application range of the DIONISIO code [11,12,13] (originally designed to predict the fuel behavior in normal operation conditions) to the high burnup domain. The predictions of DIONISIO were compared with a large number of experimental data, obtaining an excellent agreement