Administration of human GH to GH-deficient patients has yielded conflicting results concerning its impact on thyroid function, ranging from increased resting metabolic rate to induction of hypothyroidism. However, most studies have been casuistic or uncontrolled and have used pituitary-derived GH of varying purity, often contaminated with TSH. Therefore, we conducted a double blind, placebo-controlled cross-over study of the effect of 4 months of biosynthetic human GH therapy (Norditropin; 2 IU/m2.day) on thyroid function in GH-deficient adults (8 females and 14 males; mean +/- SE age, 23.8 +/- 1.2 yr). One group (I) was euthyroid without T4 substitution (n = 13), whereas the other (group II) received T4 (n = 9). Serum T4 (nanomoles per L) decreased in both groups after GH treatment [group I, 100 +/- 8 (mean +/- SE) vs. 89 +/- 8 (P less than 0.01); group II, 145 +/- 18 vs. 115 +/- 10 (P less than 0.05)]. Conversely, GH treatment caused an increase in serum T3 (nanomoles per L) in both groups [group I, 1.9 +/- 0.1 vs. 2.0 +/- 0.1 (P less than 0.1); group II, 1.7 +/- 0.1 vs. 1.9 +/- 0.1 (P less than 0.05)]. Similar changes were seen in serum free T4 and T3. The serum T3 level during the placebo period of group I was significantly lower than that in an age-matched reference group (P less than 0.02). Serum rT3 (nanomoles per L) was low in group I and decreased significantly, as in group II, after GH treatment [group I, 0.26 +/- 0.02 (placebo) vs. 0.20 +/- 0.02 (GH; P less than 0.01); group II, 0.38 +/- 0.05 (placebo) vs. 0.29 +/- 0.02 (GH; P less than 0.01)]. Serum TSH decreased in both groups during GH therapy, though not significantly. Serum thyroglobulin was unaltered and did not differ from that in the reference group. In conclusion, our data are consistent with a GH-induced enhancement of peripheral deiodination of T4 to T3. GH thus seems to play an important role, either directly or indirectly, in the regulation of peripheral T4 metabolism.