Patients with sickle-cell anemia treated with hydroxyurea may have significant reduction in frequency and severity of pain episodes. However, previous clinical trials show a variable response to hydroxyurea. Criteria which can be used to select patients who are likely to respond to hydroxyurea treatment would be useful. Our laboratory has previously demonstrated an inverse linear relationship between the total number of burst-forming unit-erythroid (BFU-E) colonies and fetal hemoglobin levels in sickle-cell patients treated with hydroxyurea. In the present report, an in vitro cell culture system was established to evaluate the effects of hydroxyurea on BFU-E colony growth and induction of fetal hemoglobin production. Five Hb SS patients who were not previously treated with hydroxyurea and three Hb SS patients who failed to respond to hydroxyurea treatment were included in the study. The results show that the number of BFU-E colonies is decreased from 153.7 to 7.2 per 3 × 105 mononuclear cells, whereas fetal hemoglobin levels were increased from 5.1 to 19.4% in the presence of hydroxyurea in vitro in cultured erythroid progenitors, which were derived from 5 patients before treatment. The number of BFU-E colonies decreased from 153.7 to 2.0 per 3 × 105 mononuclear cells in the in vitro cultures obtained from serial peripheral blood samples over a 9- to 20-week period of oral hydroxyurea therapy. A simultaneous rise in fetal hemoglobin level from 10.2 to 28.6% in the peripheral blood over the same period of hydroxyurea therapy was also observed. Our results demonstrate that the increase in fetal hemoglobin levels in cells treated with hydroxyurea in vitro is comparable to the rise of fetal hemoglobin production following hydroxyurea therapy in these patients. On the contrary, these findings were not observed in three previously non-responsive sickle-cell patients. These results suggest that the changes in number of BFU-E colonies and fetal hemoglobin levels after in vitro exposure to hydroxyurea may be a useful approach to select sickle-cell patients who will respond to hydroxyurea therapy. © 1997 Wiley-Liss, Inc.