Anemia is a common complication of critical illness. Because tissue hypoxia is a prominent factor in the development of organ dysfunction in the critically ill, conventional wisdom has argued that the transfusion of packed red blood cells can attenuate tissue hypoxia and so improve outcome.Review of pertinent English-language literature.The empiric evidence supporting the benefit of transfusion to treat tissue hypoxia is sparse; indeed, a body of recent work suggests that moderate anemia is not only well-tolerated by the critically ill patient, it is associated with improved clinical outcomes. The primary biologic rationale for transfusion of the critically ill is to maximize oxygen delivery to tissues. However, because of reflex compensatory mechanisms, and because of alterations in microvascular flow and endothelial permeability, the impact of transfusion is much less than would be predicted. Retrospective studies suggest that transfusion is immunosuppressive, and associated with an enhanced infectious risk. The large Transfusion Requirements in Critical Care (TRICC) Trial conducted by the Canadian Critical Care Trials Group demonstrated reduced mortality and organ dysfunction when a transfusion trigger of 7 g/dL is used, without an increase in infectious complications.A conservative transfusion strategy appears safe in nearly all critically ill patients without active hemorrhage, including patients with cardiovascular disease. Whether a lower transfusion threshold could be adopted is unknown.