Erythropoiesis usually fails during severe illness because of a blunting of the kidney-erythropoietin (EPO)-bone marrow axis. In this setting, clinical studies have shown that recombinant human erythropoietin (rhEPO), administered in pharmacological amounts, significantly reduces the need for blood transfusions. In addition to the kidney, however, EPO is also produced locally by other tissues in a paracrine-autocrine manner. Here, similar to its role in the bone marrow, EPO rescues cells from apoptosis. Additionally, EPO reduces inflammatory responses, restores vascular autoregulation, and promotes healing. The results of many studies (including a phase II clinical trial in ischemic stroke) demonstrate that rhEPO protects the brain, spinal cord, retina, heart, and kidney from ischemic and other types of injury. Although rhEPO is efficacious in the treatment of EPO-deficient anemia during illness, inadequate effort has been devoted to determining whether direct tissue protection might also result from its administration. Here, we speculate on the potential utility of EPO as a protective cytokine in the context of acute critical illness and suggest key parameters required for a proof-of-concept clinical study.