Aim In order to assess the influence of total-body irradiation (TBI) on the outcome and incidence of complication after bone marrow transplantation (BMT), we retrospectively analyzed our patients treated for acute leukemia and conditioned with TBI prior to BMT. Patients and Methods Between 1980 and 1993, 326 patients referred to our department with acute non-lymphoblastic leukemia (ANLL, n=182) and acute lymphoblastic leukemia (ALL, n=144) in complete remission underwent TBI either in single dose (190 patients: 10 Gy administered to the midplane, and 8 Gy to the lungs [STBI]) or in 6 fractions (136 patients: 12 Gy on 3 consecutive days, and 9 Gy to the lungs [FTBI]) before BMT. The male-to-female ratio was 204/122 (1.67), and the median age was 30 years (mean: 30 ± 11, range: 3 to 63). The patients were analyzed according to 3 instantaneous dose rate groups: 118 patients in the LOW group (≤0.048 Gy/min), 188 in the MEDIUM group (> 0.048 and ≤ 0.09 Gy/min), and 20 in the HIGH group (> 0.09 cGy/min). Conditioning chemotherapy consisted of cyclophosphamide (CY) alone in 250 patients, CY and other drugs in 54, and 22 patients were conditioned using combinations without CY. Following TBI, allogeneic and autologous BMT were realized respectively in 118 and 208 patients. Median follow-up period was 68 months (mean: 67 ± 29, range: 24 to 130 months). Results Five-year survival, LFS, RI and TRM rates were 42%, 40%, 47%, and 24%, respectively. Five-year LFS was 36% in the STBI and 45% in the FTBI group (p = 0.17). It was 36% in the LOW group, 42% in the MEDIUM group, and 30% in the HIGH group (p > 0.05). Five-year RI was 50% in STBI, 43% in FTBI, 55% in LOW, 41% in MEDIUM, and 44% in HIGH groups (STBI vs. FTBI, p = 0.48; LOW vs. MEDIUM, p = 0.03: MEDIUM vs. HIGH, p = 0.68). TRM was not influenced significantly by the different TBI techniques. When analyzing separately the influence of fractionation and the instantaneous dose rate either in ANLL or ALL patients, no difference in terms of survival and LFS was observed. Fractionation did not influence the 5-year RI both in ANLL and ALL patients. However, among the patients with ANLL, 5-year RI was significantly higher (58%) in the LOW group than the MEDIUM group (31%, p = 0.001), whereas instantaneous dose rate did not significantly influence the RI in ALL patients. The 5-year TRM rate was significantly higher in allogeneic BMT group both in ANLL (37%) and ALL (37%) patients than those treated by autologous BMT (ANLL: 15%, ALL: 18%: p = 0.002 and 0.02, respectively). The 5-year estimated interstitial pneumonitis (IP) and cataract incidence rates were 22% and 19%, respectively, in all patients. IP incidence seemed to be higher in the HIGH group (46%) than the MEDIUM (19%, p = 0.05) or LOW (25%, p = 0.15) groups. Furthermore, cataract incidence was significantly influenced by fractionation (STBI vs. FTBI, 29% vs. 9%; p = 0.003) and instantaneous dose rate (LOW vs. MEDIUM vs. HIGH, 0% vs. 27% vs. 33%; p < 0.0001). Multivariate analyses revealed that the best factors influencing the survival were 1st CR (p = 0.0007), age ≤ 40 years (p = 0.003), and BMT after 1985 (p = 0.008). The RI was influenced independently only by the remission status (p = 0.0002). On the other hand, the TRM rate was lower in patients who did not experience graft-vs.-host disease (GvHD, p < 0.0001), and in those treated after 1985 (p = 0.0005). GvHD was the only independent factor involved in the development of IP (p = 0.01). When considering the cataract incidence, the only independent factor was the instantaneous dose rate (p = 0.0008). Conclusion The outcome of BMT patients conditioned with TBI for acute leukemia was not significantly influenced by the TBI technique, and TRM seemed to be lower in patients treated after 1985. On the other hand, cataract incidence was significantly influenced by the instantaneous dose rate.