Human recombinant interleukin-3 (IL-3; Sandoz AG, Basel, Switzerland) was administered for 7 days to patients with neoplastic disease and normal hematopoiesis. The purpose of the study was to assess IL-3 toxicity, to identify target cells, to define their kinetics of response at different dose levels, and to determine if IL-3 in vivo increased the sensitivity of bone marrow (BM) progenitors to the action of other hematopoietic growth factors. A total of 21 patients entered the study; the dosage ranged from 0.25 to 10 micrograms/kg/d. The effect on peripheral blood cells during treatment showed no significant changes in the number of platelets, erythrocytes, neutrophils, or lymphocytes (and their subsets). A mild monocytosis and basophilia occurred. Eosinopenia, present in the first hours of treatment, was followed by a dose-and time-dependent eosinophilia. IL-3 treatment affected BM cell proliferation by increasing the percentage of BM progenitors engaged in the S-phase of the cell cycle. The effect was dose dependent, with the various progenitors showing different degrees of sensitivity. The most sensitive progenitors were the megakaryocyte progenitors (colony-forming unit-megakaryocyte), then the erythroid progenitors (burst-forming unit-erythroid), and finally the granulo- monocyte progenitors (colony-forming unit-granulocyte-macrophage) whose proliferative activity was stimulated at the higher doses of IL-3. Only a slight increase in the proliferative activity of myeloblasts, promyelocytes, and myelocytes was observed, whereas the activity of erythroblasts was unchanged. The priming effect was such that BM progenitors, purified from patients treated with IL-3, produced more colonies in vitro in the presence of granulocyte colony-stimulating factor (G-CSF; granulocyte colonies), IL-5 (eosinophil colonies), and granulocyte-macrophage CSF (GM-CSF; predominantly eosinophil colonies). These data indicate that even in vivo IL-3 acts essentially as a primer for the action of other cytokines. Therefore, optimum stimulus of myelopoiesis will require either endogenous or exogenous late-acting cytokines such as G-CSF, erythropoietin, GM-CSF, and IL-6 for achieving fully mature cells in peripheral blood. If exogenous cytokines are used with IL-3, it is likely that G-CSF will yield more neutrophils, whereas GM-CSF may enhance eosinophils, monocytes, and neutrophils. Attention to the clinical relevance of each cell type will be necessary and should determine the selection of the combination of cytokines.