With the electronic counters routinely used, it has become practical to determine the concentration of hemoglobin, red cell indices, and RDW concurrently in association with transferrin saturation and ferritin in accordance with feeding practices. The 1028 infants and children aged 6 to 24 months, who had been mainly admitted with acute infectious or inflammatory diseases, were divided into three groups, i.e., children who were exclusively breast-fed more than 6 months (group A), those who had been given iron-fortified formula milk since birth (group B), and those who had been given breast milk for 5-6 months and then switched to the iron-fortified formula (group C). Children with anemia comprised 34.8% (104/299) of group A, significantly more than 5.6% (34/608) of group B and 6.6% (8/121) of group C (p < 0.001, respectively). Children with MCV < 70 fl comprised 39.5% (118/299) of group A, significantly more than 7.1% (43/608) of group B and 13.2% (16/121) of group C. Out of the total 146 patients with anemia, 82.2% (n = 120) had laboratory evidence of iron deficiency, which was mostly suggested by a dietary history. The sensitivity of MCV values < 70 fl in IDA patients was 90.0%; specificity was 53.8%. The sensitivity of RDW values > or = 15% was 83.3%; specificity was 57.7%. The positive predictive value could be increased to 97.8% by combining MCV < 70 fl and RDW > or = 15%. The sensitivity of serum ferritin concentrations < 10 ng/ml was 62.4% and specificity was 100%. The sensitivity of transferrin saturation < 12% was 72.3% and specificity was 81.3%. By combining the hemoglobin with MCV and RDW in screening for iron deficiency, the diagnostic accuracy of IDA can be increased. We support the use of appropriately iron-fortified weaning foods or the routine iron supplement starting at 6 months of age in exclusively breast-fed infants.