SummaryHbA 2 , a tetramer of a-and d-globin chains, provides a diagnostic clue to the presence of b-thalassaemia trait. This minor haemoglobin, which forms about 2-3% of the total, has no known physiological role, but has the interesting property of preventing polymerization of deoxy-sickle haemoglobin. If it were possible to increase the level of HbA 2 sufficiently it could have a benefit in sickle cell disease similar to that of foetal haemoglobin. Moreover, HbA 2 is present in all erythrocytes, an advantage not found with foetal haemoglobin, which is heterocellularly expressed. The molecular basis of HbA 2 gene (HBD) expression is partially understood, and with new molecular tools, it might be possible to induce levels of HbA 2 that could be clinically important. However, high concentrations of this positively charged haemoglobin might damage the erythrocyte membrane; also, the reciprocal relationship of d-and c-globin gene (HBD and HBG1/2, respectively) expression might negate any benefit of increasing transcription of the former.