Most available clinical data strongly suggest a sparing effect of TBI fractionation for the lungs, liver, lens, the growth cartilage and, perhaps the prepubertal ovary; the usual fractionated TBI regimens, delivering from 12 to 15 Gy, appear to be constantly less toxic than the "standard" 10 Gy single dose TBI scheme. However, there is also some clinical suggestion, essentially coming from the T-depleted graft experience, that the largely used 12 Gy fractionated scheme (6 X 2 Gy) might be less effective than the standard 10 Gy single dose TBI for leukemia cell killing and for eradication of the recipient bone marrow. Additional clinical data, ideally coming from well designed randomised trial or from careful large-scale retrospective evaluations, should help to optimize the TBI delivery.