AbstractMouse bone marrow forms colonies of granulocytes and monocytic phagocytes when cultured in the presence of human plasma, urine or “feeder layers” prepared from human leukocytes. By contrast, human marrow produces colonies in the presence of leukocyte feeder layers but not in the presence of plasma or urine. It has been tacitly assumed that the response of mouse marrow to human blood leukocyte feeder layers is a measure of physiological substances released by those leukocytes which might control human granulopoiesis. This assumption however, has never been put to the test by comparing the response of mouse and human marrow to stimulation by leukocytes from the same individual. This has been done in the present study by using leukocytes from normal and leukemic subjects. Different human marrows responded similarly to stimulation by the same normal feeder layers, but there was no quantitative or qualitative correlation between the response of human and mouse marrows. Feeder layers from patients with acute granulocytic leukemia did not stimulate colony growth in normal human marrow but were as potent in stimulating mouse marrow colony growth as were feeder layers of normal leukocytes.We conclude that different factors may stimulate human and mouse marrows and that assays of granulopoietic factors of human origin should in future be carried out in human rather than mouse marr