In infants suffering from haemolytic disease of the newborn death from cardiac failure consequent upon prolonged anoxia can occur within the first twenty-four hours of life. If the infants survive this period the anoxia is relieved as blood oxygenation mechanisms become more efficient. Unfortunately, improvement is often temporary and after about forty-eight hours rapid destruction of blood cells,which commences "in utero", can result in deep jaundice. This is followed by damage to the central nervous system (known as kernicterus) with medullary failure and death in the more severe cases, Thus even though some infants are in imminent danger of cardiac failure in the immediate post-natal period, it is considered best to attempt exchange transfusion in order to reduce the risk of kernicterus. The development cf exchange transfusion has greatly altered the prognosis of haemolytic disease in the new- born and the procedure Is now accepted as being essential in cases of any severity. The difficulties associated with blood transfusion have received increasing attention in recent years and its employment Is beset with problems not only of cytology, but of micro-biology, biochemistry, allergy, dynamics, religion and race, Some of these problems are significantly affected by the volume of blood use in exchange transfusion. Considering that the infant may be perfused with stored blood of electrolyte content greatly different from that or the recipient's blood and in quantities up to three times his own blood volume, it is surprising that so little clinical change Is observed. Hazards such as sepsis, air embolism, portal vein thrombosis and perforation of the umbilical vein have peen described but are unfortunately uncommon. Circulatory overloading is an ever present danger in the profoundly anaemic infant and it may be precipitated by conducting the exchange transfusion too rapidly. Deterioration and even death however, may occur in infants whose preoperative condition gives rise to no concern, An example of such deterioration prompted this investigation. During the exchange transfusion of a male infant, the infant suddenly collapsed and appeared pale and unresponsive, calcium gluconate did not affect an Improvement and administration of glucose only caused a temporary improvement. The transfusion was carried to a satisfactory conclusion In negative balance of blood. Previous work on citrate intoxication was recalled and it was decided to undertake a combined clinical and biochemical study of infants during exchange transfusion.