The hematopoietic system in patients with aplastic anemia (AA) shows both quantitative and qualitative deficiencies, i.e., reduced numbers of hematopoietic progenitor cells (HPC) and impaired HPC proliferation in long-term marrow cultures (LTMC). Since recombinant human granulocyte macrophage-colony stimulating factor (rhGM-CSF) has been shown to be a potent stimulator of normal hematopoiesis, both in vivo and in vitro, in the present study we wanted to assess the possibility of stimulating hematopoiesis in LTMC from 17 patients with AA, by weekly addition of rhGM-CSF (10 ng/ml). In LTMC from 11 patients (group of responders), rhGM-CSF induced a significant increase (4.8-fold, compared with untreated cultures) in the levels of myeloid progenitor cells; in contrast, in six patients (group of nonresponders), myeloid progenitors were refractory to this cytokine. In the group of responders, rhGM-CSF also induced a pronounced increment in the levels of nonadherent and adherent cells (5.99-and 5.18-fold, respectively, compared with untreated cultures). Among the different myelopoietic lineages, rhGM-CSF preferentially stimulated the macrophagic lineage; this was evident both at the progenitor and mature cell levels. Interestingly, the effect of rhGM-CSF in LTMC from AA patients was only transient. Indeed, the effects mentioned above were observed only during the first three weeks of culture; afterwards, myeloid progenitor and nonadherent cell levels in treated cultures declined, practically reaching the levels observed in untreated cultures. At the moment, we do not know whether this transient stimulatory effect is due to the production of inhibitory cytokines, by macrophages generated in response to rhGM-CSF, or to the exhaustion of the HPC pool in AA cultures. In all 17 patients, rhGM-CSF had no effect on the kinetics of erythroid or multipotent progenitor cells. These results are in keeping with clinical studies in which it has been observed that most AA patients treated with rhGM-CSF show increments in circulating monocytes and granulocytes, as well as in bone marrow cellularity. However, little or no effect is observed on erythropoiesis. The actual mechanisms involved in the in vitro effects of rhGM-CSF on myeloid progenitor cells from AA bone marrow are still not completely understood. Future studies on this issue should be encouraged, since they may help to understand the in vivo (clinical) effects of this cytokine. Am.