A hypothesis of the pathogenesis of blackwater fever, suggested by the observation that liver emulsion from a rhesus monkey dead of yellow fever, stimulates the formation of antiliver antibodies when injected into another rhesus monkey, is put forward and discussed. This hypothesis may be formulated thus:— ▪ ▪ It is noted that, in cases of blackwater fever, the red cells show auto-agglutination and spherocytosis and that erythrophagocytosis is commonly seen in smears of the spleen, and, occasionally, in films of peripheral blood: Further, it has been observed, red cells from healthy donors haemolyze as rapidly in a case of blackwater fever as the patient's own cells and, on the other hand, red cells from a case of blackwater fever, transfused into a healthy recipient, haemolyze as rapidly as the red cells remaining in the patient. There is usually a considerable interval between the first attack of malaria and the onset of blackwater fever, during which the patient suffers from repeated attacks of malaria, often inadequately treated with quinine. It is noted, too, that blackwater fever is closely simulated by the effects of an incompatible blood transfusion by paroxysmal haemoglobinuria and by the effects of the injection of an artificially produced haemolytic serum into experimental animals. All these observations are readily explained on the assumption that a biological autohaemolysin is concerned in the pathogenesis of blackwater fever. Other hypotheses put forward cannot satisfactorily account for them, especially for the specific association of blackwater fever with malaria. It is concluded, therefore, that an autohaemolysin is almost certainly concerned in the production of blackwater fever. The site of formation of this autohaemolysin is discussed, and it is considered that it is mostly produced in the spleen. The boosting of the titre of the autohaemolysin resulting from repeated attacks of malaria often inadequately treated, and factors producing a sudden contraction of a congested spleen, in which autohaemolysin has accumulated, are considered likely to precipitate an attack and to account for the sudden dramatic onset of blackwater fever. The treatment of the condition in the light of this hypothesis is discussed, and it is considered important to prevent such sudden contractions of the spleen. It is also important, in order to avoid the development of a state of sensitivity, to treat malarial infections adequately. Other conditions characterized by intravascular haemolysis are briefly discussed. It is noted that apart from incompatible blood transfusion, and paroxysmal haemoglobinuria which are known to be caused by biological antibodies, such conditions as the acute haemolytic anaemias following certain drugs and of favism can also be reasonably explained. Cases of aplastic anaemia, of agranulocytosis, and of thrombocytopenia, are briefly described, and it is noted that an explanation for their development is possibly to be found in the hypothesis here enunciated.