BpIjryLo 1647 haemorrhagic diathesis as a cause of the bleeding (Jacobson, 1941).Even if we could not show any effect on the haemostasis in normals, premarin might well act as a haemostatic agent clinically, through some as yet unknown mechanism. We chose to test this out in epistaxis and in surgery on the prostatic gland, two conditions in wljh the effect of premarin has been reported to be excellent.We could not confirm the results of Bleeker (1959), who found that in prostatectomy the number of blood transfusions could be reduced by 50% when premarin was used. There can be no doubt that if premarin is given in surgery the evaluation of the effect should be done with a blind technique. If this cannot be done, it must be preferable to measure the actual loss of blood as in this study, rather than to rely on changes in haemoglobin values or impressions of the blood loss. There was no difference in blood loss between patients who received premarin, either preoperatively only or both pre-and post-operatively, and those who did not receive premarin. Our findings are in good agreement with the result of the double-blind study of Cooner and Burros (1960).In this connexion we would like to point out that, according to Jacobson, "spontaneous bleeding" excludes surgical and traumatic bleeding. Therefore there should be no indication for premarin therapy in such cases.Jacobson (1954) states that " spontaneous bleeding" is often preceded by tension or stress. One may assume then that the bleeding in itself will increase the anxiety and a vicious cycle will then be started. All kinds of treatment which would tend to break this cycle somewhere will reduce the actual bleeding. This should be kept in mind when the effect of premarin in epistaxis is evaluated. When a physician takes care of such a patient, talks with him, puts him to bed, gives him any kind of medicine, he will calm the patient and tend to break the cycle. This treatment in itself would be a major factor in controlling the bleeding. If premarin has any effect over and above this, it must be proved in a controlled study with a double-blind technique. This has never been done, and the clinical reports on premarin are therefore of little value in our opinion. Our double-blind study is unfortunately small, but it does suggest that the effect of premarin is only that of a placebo.The theory that " spontaneous bleeding" is a clinical entity which is due to a low oestrogen level in the blood seems to us to be very speculative and loosely founded. So far as we know, this postulated low level of oestrogen has never been demonstrated. If the theory be correct, it is difficult to understand why there is not a striking difference in the incidence of spontaneous bleeding between men and women.We have confirmed that premarin produced no sideeffects. We think, however, that this fact in itself is no justification for the extensive use of premarin. We are afraid that such a drug will give the doctor who is using it a false sense of security, and that this, together with the ...