Rats were fed propylthiouracil (0.05 %), methimazole (0.05 %), or sulfaguanidine (1 %) and injected with graded doses of thyroxine (0.5-3.0 jug) daily for 2 weeks. In spite of the high serum PBI with doses of 1.0 and 1.5 jug of thyroxine, goiter was still observed in animals fed propylthiouracil. Goiter was prevented only when the PBI was increased to 2.2 times that of the control level by giving 2.0 jug of thyroxine. Similar effects were found with methimazole and sulfaguanidine. In animals which were thyroidectomized and given thyroxine, methimazole slightly elevated the serum PBI and reduced the urinary excretion of radioiodine after a single injection of labeled thyroxine. However, since sulfaguanidine was without effect in these respects, the development of goiter in the presence of a high serum PBI cannot be explained by assuming that all 3 goitrogens interfered with the deiodination of thyx'oxine. On the other hand, goiter was prevented in propylthiouracil-fed animals by doses of thyroid powder which resulted in a serum PBI lower than normal. Furthermore, goiter prevention in methimazolefed animals was achieved by administering a mixture of triiodothyronine (0.1 jug) and thyroxine (1.0 /xg) which produced a normal serum PBI and thyrotropin level. Since normally triiodothyronine plays an important physiologic role, it is suggested that goiter could not be prevented without giving a dose of thyroxine that would elevate the serum PBI to compensate for the activity of triiodothyronine. The existence of a homeostatic relation between the level of serum PBI and thyrotropin secretion by the anterior pituitary can be shown in goitrogentreated animals only when an appropriate mixture of triiodothyronine and thyroxine is administered. (Endocrinology 82: 91, 1968) AS FIRST formulated by Aron et al.(1) J\. and then reiterated by Hoskins (2), it is generally held that the pituitarythyroid axis offers an example of a physiologic servomechanism. When the titer of circulating thyroxine rises, the anterior pituitary is selectively inhibited, and thereby the output of thyrotropin falls. On the other hand, episodic or persistent decrease of circulating thyroxine results in an increased secretion of thyrotropin. Since probably over 90% of the iodine in blood is represented by thyroxine (3-5), it has been suggested that the level of thyroxine, measured as PBI, may regulate the secretion of thyrotropin. Although most