IL-5 is induced locally in the lung and systemically in the circulation during allergic airways eosinophilic inflammation both in humans and experimental animals. However, the precise role of local and systemic IL-5 in the development of allergic airways eosinophilia remains to be elucidated. In our current study, we demonstrate that compared with their IL-5(+/+) counterparts, IL-5(-/-) mice lacked an IL-5 response both in the lung and peripheral blood, yet they released similar amounts of IL-4, eotaxin, and MIP-1alpha in the lung after ovalbumin (OVA) sensitization and challenge. At cellular levels, these mice failed to develop peripheral blood and airways eosinophilia while the responses of lymphocytes, neutrophils, and macrophages remained similar to those in IL-5(+/+) mice. To dissect the relative role of local and systemic IL-5 in this model, we constructed a gene transfer vector expressing murine IL-5. Intramuscular IL-5 gene transfer to OVA-sensitized IL-5(-/-) mice led to raised levels of IL-5 compartmentalized to the circulation and completely reconstituted airways eosinophilia upon OVA challenge, which was associated with reconstitution of eosinophilia in the bone marrow and peripheral blood. Significant airways eosinophilia was observed for at least 7 d in these mice. In contrast, intranasal IL-5 gene transfer, when rendered to give rise to a significant but compartmentalized level of transgene protein IL-5 in the lung, was unable to reconstitute airways eosinophilia in OVA-sensitized IL-5(-/-) mice upon OVA-challenge, which was associated with a lack of eosinophilic responses in bone marrow and peripheral blood. Our findings thus provide unequivocal evidence that circulating but not local lung IL-5 is critically required for the development of allergic airways eosinophilia. These findings also provide the rationale for developing strategies to target circulating IL-5 and/or its receptors in bone marrow to effectively control asthmatic airways eosinophilia.