Granulocyte colony-stimulating factor was discovered during attempts to define the normal regulators present in cell supernatants that could induce terminal differentiation of the murine myeloid leukemic cell line WEHI-3B D+. The purification and subsequent cloning of both murine and human G-CSF allowed the normal functions of this molecule to be elucidated and indicated that it was a relatively lineage-specific stimulator of the survival, proliferation, and differentiation of precursor cells of the neutrophilic granulocyte cell lineage as well as an activator of mature neutrophil function. Murine and human G-CSFs as well as the ligand-binding domains of G-CSF receptors have been strongly conserved so that the biological activities and receptor-binding characteristics of G-CSFs are completely species cross-reactive. The evolutionary conservation of G-CSFs is suggestive of an important role for the molecule in maintaining neutrophil levels and activity in vivo, a role amply supported by a series of animal experiments and clinical trials that have been performed recently. These experiments and trials have suggested that G-CSF will be clinically useful in augmenting patients' resistance to certain infections, in enhancing the neutrophil responses of some patients with reduced granulocyte counts or defective granulocytes, and in reducing the decline in granulocytes that occurs during chemotherapy or irradiation therapy and bone marrow transplantation. The special capacity of G-CSF among the CSFs to induce terminal differentiation in some myeloid leukemias is not understood in molecular terms. However, the capacity of G-CSF to stimulate proliferation in some myeloid leukemias as well argues that caution needs to be exercised in the timing of G-CSF administration and patient selection before considering this form of therapeutic approach.