The behaviour of a 48Y cylinder subjected to a regulatory fire of 800uC for 30 min, in accordance with the criteria from the International Atomic Energy Agency, has been the subject of controversy for several years. Some claimed that if the 48Y cylinder is filled to the nominal value of 12. 5 tonnes of UF 6 and subjected to a 30 min fire, it would fail, according to an interpretation of the results from the TENERIFE experimental programme and a preliminary model. Others who did not have access to the TENERIFE results, claimed that the 48Y would not fail. A more in depth study showed that the preliminary model is not able to ascertain either the resistance or the failure of the 48Y cylinders subjected to the 30 min fire. This led the authors to revisit all the available experimental data and more particularly the results from the TENERIFE experimental programme and to look with particular care at the temperature inside the cylinder and to the pressure evolution after cessation of heating. It was concluded first that the temperatures inside the TENERIFE cylinder at the end of the heating period allow the preliminary model to be adjusted, and second that the simulation of the experimental results after approximately 2000 s by the preliminary model, was only possible by using non-physical and arbitrary numerical operations. Finally, a more realistic mechanism for the increase in pressure before and after cessation of heating is suggested.