THE CHARACTER and significance of epistaxis have never been understood. In 1941 I proposed the hypothesis that spontaneous hemorrhage is a clinical entity in itself and that epistaxis is the commonest form of this disease.1 Some time later an elaboration of this concept, with a report of 49 patients who had 73 bleeding episodes, was published.2 Since then this study has continued and broadened into other fields, but no similar experience appeared until 1952, when Koch and his associates3 reported five patients with hereditary hemorrhagic telangiectasia relieved by the same means as I had employed. Although I have had the experience of successfully treating well over 300 patients with nosebleeding by the intramuscular administration of estrogens, this contribution contains the important features of the cases of only eight patients with overwhelming epistaxis controlled by huge doses of estrogens administered intravenously. Four other cases of peculiar interest will