Fifty patients with total joint arthroplasties (28 total hip arthroplasties, 11 total knee arthroplasties, and 11 bilateral total knee arthroplasties) received autotransfusions from their postoperative wound drainage. The blood was collected in a closed sterile drainage system without any additional anticoagulant. Pre- and postoperative measurements were made of the patient's hemoglobin, platelets, fibrinogen, haptoglobin, fibrin degradation products, and D-dimer (a specific type of fibrin degradation product). Red blood cell survival was assessed in 16 of the patients by labeling the shed blood with 51Cr sodium chromate prior to reinfusion. To control for fluid shifts, continued bleeding, and dilution effects of further transfusions in the immediate post-operative period, 10 patients also had their native blood labeled with 111In oxime. In this study, the mean estimated blood loss was 1,062 mL (±1,247) with a mean wound drainage of 836 mL (±338). Of this, a mean of 450 mL (±261) of blood was was given back to the patient in addition to routine, preoperative autologous donated blood. Six (12%) patients experienced transient fevers at the time of retransfusion. Detailed hematologic studies were performed on the shed blood in 19 patients. The collected blood was completely defibrinated, but did contain fibrin degradation products, as indicated by the D-dimer level, and hemolyzed blood as the haptoglobin was reduced. Even though the blood containing the above breakdown products was reinfused to the patients, there were no clinical manifestations of disseminated intravascular coagulation. Both the hemolyzed and defibrinated products were subsequently cleared by the body. After correcting for fluid and volume changes in the first 3 days, 66% (SEM, 5.3%) of the labeled red blood cell activity remained and was used to determine a mean red blood cell survival half-life of 40.6 days (SEM, 4.2). There was no significant change in the 111In: 51Cr: ratio in the first 24 hours, indicating that the initial drop in activity equally affected both the native red blood cell population and the labeled shed blood. There were no adverse clinical effects resulting from a transient coagulopathy. Autotransfusion of postoperative shed red blood cells is a safe method with reinfused cells that have a normal or increased lifespan.