Vomiting of blood is one of the commoner medical emergencies and is consequently a condition with which all physicians have fairly frequently to deal. Much has been written about it and there is still a good deal of disagreement, especially regarding its treatment. It is proposed in this paper to discuss its aetiology, prognosis and treatment and to analyse a collection of cases. The cases to be considered are all those admitted because of haematemesis, to the Leicester Royal Infirmary between the beginning of 1432 and the end of February 1937. Only those patients will included who were admitted because of recent haematemesis and not those who had melaena only, or those who gave a history of having vomited blood at some time, but who were admitted for other reasons. Cases complaining of melaena have been excluded because it is difficult to judge the severity of their bleeding. 1. 188 cases of haematemesis have been analysed. 2. It is concluded that about 13 of cases are due to cirrhosis of the liver, splenic anaemia, carcinoma of the stomach, and rare causes. 3. The other 90% are due to (a) Chronic peptic ulcer. (b) Gastritis (or gastroduodenitis) and these two causes are probably equally common. The gastritis may be acute or chronic. 4. It seems likely that many patients who have a haematemesis possess some peculiarity which makes them prone to bleed. 5. Haematemesis is very liable to take place after gastroenterostomy, especially if the patient bled before the operation. 6. Haematemesis is more common in men than in women. 7. The average age of patients suffering from haematemesis is about 40. 8. In cases of a severity requiring admission to hospital the immediate mortality in haematemesis from all causes is about 15 /, or possibly more: in haematemesis from peptic ulcer and gastritis is not less than 10. 9. In the series of cases analysed, half of the patients took 6 months or more to get back their normal strength, very few of them regained perfect health and about half continued to have a significant degree of indigestion in spite of diet. 10. It is suggested that lack of food and drink may be a factor sometimes in causing death. Fluids were given by mouth to a few of the cases in this series, starting a few hours after admission. This method of treatment might be tried out further. 11. If this treatment is adopted it is probably wiser not to give rectal salines. 12. Blood transfusion should usually be carried out if the haemoglobin is under 305 and occasionally when it is over 30, the determining factor being the clinical condition. 13. Immediate operation is of no value except in cases of chronic ulcer, and in them its value is doubtful. Further evidence regarding treatment by operation combined with drip transfusion is desirable.