(Footnotes continued on next page) Introduction Administration of oxygen to premature infants had been implicated as an etiologic agent in the development of retrolental fibroplasia (RLF) prior to late 1952.* But because of the apparently capricious fluctuation in incidence of the disease observed in various communities and the insufficiency of controlled investigations concerning the possible role of oxygen in RLF the evidence was deemed inadequate to justify curtailment of oxygen for premature infants, particularly in the absence of knowledge of the effect on mortality of such a change in procedure. For these reasons a number of pediatricians and ophthalmologists agreed early in 1953 to pool their nursery facilities and interests and investigate cooperatively the question of whether oxygen was in fact an etiologic agent in RLF and at the same time to determine what effect restricting oxygen might have on infant mortality. This group of investigators, from 18 hospitals located