Fetal cells in maternal peripheral blood are a source of fetal DNA for prenatal genetic diagnosis, but their numbers are so small and variable that a reliable isolation procedure has yet to be demonstrated. The problem of scarcity may be overcome by amplification of fetal progenitor cells in cultures from maternal blood samples. One challenge is to identify post-culture fetal cells and colonies. We have found that the progeny of fetal and adult erythroid progenitors developed differential haemoglobin profiles in co-culture. Fetus-derived cells initially made only fetal haemoglobin (HbF) and began to express adult haemoglobin (HbA) only after intracellular HbF had reached maximum levels, which occurred after c 7 d in culture. By this time the large majority of adult-derived erythroid cells contained already high levels of HbA alone or combined with HbF. Using the HbF+ HbA- criterion, we were able to flow-sort fetal cells with up to nearly 50% purify from some post-termination blood cultures, and with >90% purity in cultures from maternal blood spiked with 1% blood from the fetus. Fetal cell purity depended on culture time and serum supplement. After 7-10 d, purity was higher in low concentrations of human cord serum (1-3%) than in the standard 30% fetal calf serum. This was reversed at later times. Thus, if fetal clonogenic erythroid cells were present in maternal blood, their progeny could be isolated from most adult erythroid cells based on haemoglobin profiles. Cultures using CD34+ cells could be performed complementary to other methods targeting more mature fetal cells in the same maternal blood samples, thus increasing the overall chances of finding fetal cells and potentially providing clonal isolation of such cells.