At the Institut Gustave-Roussy (IGR), from January 1982 to December 1986, 54 patients received total body irradiation (TBI) as a part of the conditioning regimen before allogeneic bone marrow transplantation. The patients were non-randomly assigned to either single dose TBI (STBI) (31 patients receiving 10 Gy at a 4.5 cGy/min dose rate, 8 Gy to the lungs) or to a hyperfractionated scheme (HTBI) (23 patients receiving 13.2 Gy in 11 fractions, 3 fractions per day, 9 Gy to the lungs). Relapse rate and overall survival were not significantly different in the two STBI and HTBI groups, in spite of a larger number of 2nd and 3rd remission patients in the HTBI subset. The incidence of interstitial pneumonitis (IP) was significantly reduced in the HTBI group (13%, versus 45% after STBI, p = 0.02). Lethality by IP was also lower after HTBI (4%, versus 26% after STBI, p = 0.08). There was no case of veno-occlusive disease of the liver in the HTBI group, whereas three cases were observed after STBI. Based on these results, the IGR activated, in January 1987, a randomized trial comparing the single dose 10 Gy TBI (8 Gy to the lung) to a new hyperfractionated schedule (11 fractions of 1.35 Gy, 3 fractions per day, 9 Gy to the lungs).