Although it is almost certain that a + -thalassemia protects against malaria, the mechanisms for that are still unknown. It has been suggested that an increased number of young circulating red blood cells in a + -thalassemic children, as a result of some degree of ineffective erythropoiesis, could be related to the high frequencies of the a + -thalassemic allele in malaria endemic areas. Reticulocyte evaluation in this condition, however, has been poorly performed so far. Our objective was to determine the reticulocyte number and maturation degree, in addition to the soluble transferrin receptor and serum erythropoietin levels, in a + -thalassemia heterozygotes, comparing them with normal a-genotype controls. One hundred twenty-one a + -thalassemia carriers (-a 3.7 /aa) and 249 controls (aa/aa), all of them with normal serum ferritin levels, were subclassified according to age (1-5, 6-10, 11-15, 16-20, and over 20 years old). Reticulocyte analyzes were carried out by flow cytometry and sTfR and s-Epo levels determined by immunonephelometry and chemiluminescence, respectively. The comparisons did not show any significant difference between thalassemics and controls regarding the reticulocyte parameters [percentages and absolute values, P = 0.2643 and 0.5421; high, medium, and low maturation degree, P = 0.2579, 0.2196, and 0.4192; RET maturity index (RMI), P = 0.2471, respectively], as well as the s-Epo levels (P = 0.5711). The sTfR concentrations were higher in the thalassemic group (P = 0.0001), but statistical significance was due only to the 1-5 and over 20 subgroups (P = 0.0082 and 0.0436, respectively). The results found here are compatible with a compensated erythropoiesis and do not confirm the hypothesis mentioned above. Am.