The authors compare the relative validity of HbA1c and fructosamine as indices of long-term blood glucose levels in population with a common occurrence of haemoglobinopathies. HbA1c levels in patients with haemoglobinopathies do not reflect the glucidic status of patients when compared to HbA1c levels in reference population. Validity of the fructosamine assay allows its potential use as a mass screening test for diabetes in these populations (USA, Africa, Caribbean...).