1. It is widely believed that a large fraction of the blood granulocyte population is located in the pulmonary circulation. 2. Evidence in favour of this belief is based on several independent and complementary techniques including studies of granulocyte deformability in relation to pulmonary capillary diameter, isolated perfused lungs, direct videomicroscopic observations, cellular and capillary morphometry, physiological studies and labelled blood cell kinetics. 3. Inconsistencies in this body of evidence can be identified and traced in many cases to doubts concerning the physiological integrity of labelled granulocytes. 4. In addition to the lungs, other regions of the body undoubtedly pool granulocytes, and there is good quantitative evidence, based mainly on clinical studies, pointing to the liver, bone marrow and especially the spleen. 5. Clinical studies, furthermore, have generally not supported major granulocyte pooling in the lung, except in association with systemic inflammatory diseases, such as inflammatory bowel disease and systemic vasculitis. 6. Compromises based on technical considerations permit reconciliation of the literature, and an overall whole-body model of granulocyte distribution is proposed with which all the existing data are broadly compatible.