Some unsolved problems--late onset of anemia and growth retardation (at age 7 years), healthy siblings showing very low transferrin (TF) level, and unexplained mode of inheritance--were found in family members of a congenital atransferrinemia already reported in 1972. The long-term clinical, laboratory, and developmental observations revealed that after 5 years of apo-TF supplementary therapy the patient's anemia gradually disappeared, and he started to grow again without further therapy. Immunoelectrophoretic study disclosed a severe deficiency of both TF and haptoglobin in the patient. The recovery from his anemia and the resumption of his physical development were dependent only on his TF level: that is, from a negligible level it increased to 10-20 mg/dl (normal, 205-370 mg/dl), a level similar to that of his TF-deficient siblings, who had been in good health since birth. The TF analysis of the patient and his family suggests that the minimum TF requisite in this family may be close to 10-20 mg/dl; subjects with more than 20 mg/dl are apparently healthy; with less than 10 mg/dl they may develop severe growth retardation and anemia, and extreme deficiency may be lethal. Also, coexisting haptoglobin deficiency might alleviate hemosiderosis. Further, the isoelectric focusing study disclosed that there was only a small amount of TF variant in these siblings including the patient. The study of the family confirmed that this variant was produced by an allelic gene derived from their father. So, the original diagnosis of congenital atransferrinemia should be revised as familial hypotransferrinemia transmitted with autosomal recessive mode, and the subjects with a recessive character may be compound heterozygotes of the "variant" allele and the "null" allele.