The sensitivity of healthy tissue is the limit of radiation therapy. Most of the technological progress of radiotherapy along its history have been devoted to improve the sparing of healthy tissues while increasing the tumor dose. The use of hadrons has been justified first for neutrons by radiobiological considerations (high RBE and no oxygen effect), then for protons by ballistic considerations (Bragg peak). Neutron ballistic is so unfavourable and side effects are so important when large and deep tumors are irradiated that their indications have been dramatically reduced and no recent studies on late effect after high RBE irradiation have been carried out. Therefore, the renewal use, hopefully on a large scale, of high RBE beam such as carbon ions, demands a new appraisal of the radiation tolerance of healthy tissues. From the photon and the neutron experience should be drawn: (i) the typology of late effect; (ii) the time schedule of the observation, (iii) the equivalent dose associated with observable late effects; (iv) dose limitations according to the type of tissues, the acceptable adverse effects, the security required and the RBE of the part of the beam irradiating the healthy tissues. Commune scale and description of the late effects will have to be decided. The biological mechanisms of late effects associating slow cellular lost, micro vascular failure and active fibrosis will need further investigation in the frame of high RBE irradiation.