T HE maintenance of the fluid balance of surgica1 patients is one of the most important responsibilities of the modern surgica1 team, and the proper repIacement of bIood is perhaps one of the most important phases of this maintenance. This is true not only because it is an excellent prophyIactic measure against shock but also because it aids in decreasing postoperative morbidity. In order to obtain optimal effects, however, the blood must be replaced at the proper time-.as the patient loses it-and in the proper amount. AIthough most surgeons and anesthesiologists we11 appreciate this fact, they are often faced with the probIem of accurately estimating bIood loss in order to avoid under-replacement or over-replacement, both of which are extremely important in poor risk patients. Purely subjective estimation by the surgical team has been shown to be often grossly inaccurate and at times the errors of personal judgment may reach dangerous levels, particularly in the poor risk patient. In order to solve this problem several methodsle5 have been proposed, but of these the direct measurement as the operation progresses is the onIy practical and reliable means of determining blood loss. It is the purpose of this articIe to present a simple and practical method which we have used to accompIish this in 748 operations and also to present a complete review of the American literature. before surgery. They commented on the Iarge Ioss of bIood during certain operations, particuIarIy radical mastectomy in which there seemingIy was no excessive bleeding at any time during the procedure. BIain’ in 1929 discussing his experience with 3,000 transfusions emphasized the fact that the amount of bIood lost during operations is often several times greater than estimated by the surgeon. He urged the preoperative correction of anemia and the immediate repIacement of bIood Ioss during operation and condemned the procrastination of some surgeons and the deIay in giving bIood transfusions unti1 after shock had deveIoped. GoIIer and Maddocks in 1932 reported their observations in measuring IIuid Ioss incIuding bIood during eighteen operations and concluded that the amount Iost was greater than they had estimated. In 1937 Pilcher and Sheardg reported the average bIood loss in two series of transurethral prostatic resections as compared to a group of forty-nine genera1 surgical cases. The first series of fifty&e cases had a blood Ioss of 479 cc. and in the second series studied after aIterations in their technic the blood Ioss was reduced to 291 cc. HubIy’O in the same year, while investigating the hemostatic properties of congo red during this type of operation, determined the postoperative bIood loss was at Ieast two-thirds of that which occurred during operation. The measurement of blood Ioss during surgiWhite et al.” studied bIood Ioss in thirtycal operations has been reported by several seven neurosurgical operations and found that authors and in some clinics it has become a the Ioss during extensive intracrania1 proceroutine procedure, particularly in poor risk dures averaged from 500 to 1,200 cc. and even patients. Gatch and LittIe6 in 1924 were the up to 2,000 cc., expIaining such reIativeIy Iarge first to report measurement during some of the loss as due to extreme vascuIarity of the scalp more common operations in genera1 surgery. and diffrcuIty in controlling the bleeding in They used the acid hematin method which inthese cases. They aIso determined that the voIves washing a11 the sponges, linen and instrureduction of concentration of red bIood cells, ments free of bIood and then adding hydrohemogIobin and hematocrit readings continues chloric acid to make the washing a 0.1 N progressively and is found to be Iowest during solution prepared from the blood of the patient the fourth postoperative day. In 1938 Stewart * From the Departments of Anesthesia and Surgery, Tacoma General HospitaI and Madigan General HospitaI, Tacoma, Wash.