BACKGROUND: Interpreting the erythroid lineage in populations with high frequency of α+ thalassemia allele is challenging due to the high prevalence of α+ thalassemia homozygotes. For such populations, separate reference values for normal and α+ thalassemia homozygotes are needed. METHODS: We studied the erythroid lineage in 1,079 citizens of United Arab Emirates (UAE). Subjects with abnormal hemoglobin (39), iron deficiency (136) or erroneous entries (8) were excluded. MCV distribution in the remaining individuals (896) was visibly bimodal. Statistical mixture analysis with Normix program was used to separate subpopulations with normal and small red cells. Hardy-Weinberg equation was used to estimate genotype frequencies. RESULTS: MCV of 78.0 fl separated phenotype-derived normal homozygotes (715) from phenotype-derived α+ thalassemia homozygotes (181). The erythrocyte indices were significantly different between the two groups (p < 0.0001). The overall prevalence of phenotype-derived α+ thalassemia homozygotes (-α/-α) was 0.20 and markedly varied among tribes, 0 to 0.31 (Mean = 0.15). The frequency of phenotype-derived α+ thalassemia allele was 0.44; when accounting for tribal population structure and inbreeding, the calculated frequency was 0.34. These values were very similar to those found in the same population by genotyping and other phenotyping methods. The erythrocyte reference values for phenotype-derived normal homozygotes in Emiratis closely overlapped with those for Caucasians and normal homozygotes defined by genotyping. The reference values for phenotype-derived α+ thalassemia homozygotes in Emiratis also closely overlapped with those for α+ thalassemia homozygotes defined by genotyping. CONCLUSION: In populations with frequent α+ thalassemia mutations, two sets of erythrocyte reference values could be determined without genotyping.