It is shown that the concepts evolved by Jacob & Monod (1961a,b) as an outcome of their studies on the lactose system in Escherichia coli can be applied to a mammalian system, namely to the synthesis of human hemoglobin polypeptide chains, in a way that appears consistent with the known facts. In the system involving the β, δ- and γ-chain genes the intervention of two distinct operators is postulated. Reasons are given why the regulator gene involved in the control of γ-chain production is probably not identical with any of the known structural hemoglobin chain genes. The activity of the γ-chain gene is considered to be submitted to a double control, namely on the one hand to an interplay of regulator genes and operators, and on the other hand to an inducer that reacts with the product of a regulator gene. Physiological conditions are thought to influence the state of activity of the γ-chain gene via the inducer. Familial fetal hemoglobinemia is considered as an operator-negative mutation. Other mutations are examined that would lead to a phenotypically comparable expression. The occurrence of some as yet unobserved mutations, such as familial adult hemoglobinemia in the fetus, is predicted. In opposition to suggestions that have been made, β-thalassemia, in typical cases, appears to be attributable neither to a mutation in the affected structural gene, nor to a mutation in an operator-like factor. It might be attributable to a translocation (or inversion), or to a “silent mutation” in an unknown structural gene located between the β-chain gene and its operator. The probable role of controller-gene diseases in evolution is pointed out.