Infants <1500g birth-weight, and expected to survive >48 hours were randomly assigned to oral ∞-tocopheral 25mg (E) or drug vehicle (C) within 2 prognostic strata (B. Wt., and disease severity) in 4-patient blocks. Sample-size was determined using ∞=.05, β=0.10 for 50% reduction in incidence of BPD from an estimated 32%. Stage III and IV BPD was diagnosed by 3 independent radiologists on chest x-ray at 3, 4 and 6 weeks. Serum E levels at 1 week were significantly different (E = 2.83 ± 0.29, C = 0.89 ± 0.08) suggesting adequate absorption and compliance. Results on the first 138 patients wereBPD was not prevented by E. The groups were similar in B.Wt., G.A., sex-ratio, and ventilator and 02 exposure in 1st 2 weeks. BPD occurred in 13/24 neonates receiving E, with B.Wts. <1000g and 13/30 controls <1000g. (N.S.). It was confined to ventilated infants; 13/20 E infants <1000g and ventilated, developed BPD compared with 13/23 similar controls. The results are unchanged if deaths in the first 2 weeks are included as BPD.