alpha-Subunit and beta-subunit of TSH were measured in the sera of five patients with idiopathic central hypothyroidism due to the secretion of biologically inactive TSH, in seven normal controls matched for bone age and sex, and in five subjects with mild primary thyroid failure before and after TRH (200 micrograms, iv) stimulation. Basal serum alpha-subunit concentration in patients did not differ from that in normal controls (mean +/- SD, 0.40 +/- 0.20 vs. 0.38 +/- 0.28 ng/ml; P, NS), whereas TSH and TSH-beta were significantly higher in patients (TSH, 1.51 +/- 0.74 vs. 0.59 +/- 0.53 ng/ml, P less than 0.025; TSH-beta, 0.56 +/- 0.18 vs. 0.10 +/- 0.02 ng/ml, P less than 0.001). The concentration of TSH-beta was also significantly higher in patients with central hypothyroidism than in subjects with mild primary thyroid failure (0.56 +/- 0.18 vs. 0.24 +/- 0.08 ng/ml; P less than 0.01), although serum TSH levels did not differ in the two groups (1.51 +/- 0.74 vs. 2.16 +/- 0.52 ng/ml; P, NS). alpha-Subunit was significantly higher in primary hypothyroid subjects (1.50 +/- 0.87, P less than 0.05 compared with patients with central hypothyroidism). After TRH, alpha-subunit, TSH, and TSH-beta net increases (peak) were significantly higher in patients with central hypothyroidism than in normal controls (alpha-subunit: 0.95 +/- 0.5 vs. 0.47 +/- 0.19 ng/ml, P less than 0.05; TSH: 7.1 +/- 3.1 vs. 2.9 +/- 1.8 ng/ml, P less than 0.005; TSH-beta: 0.89 +/- 0.35 vs. 0.22 +/- 0.18 ng/ml, P less than 0.005), whereas they did not significantly differ from those recorded in hypothyroid controls. The beta/alpha ratio, which was 1.67 +/- 0.86 in patients and 0.35 +/- 0.18 in normal controls (P less than 0.005), slightly decreased after TRH to 1.24 +/- 0.78 in patients, but remained unchanged in normal controls (0.39 +/- 0.1). After TRH the alpha-subunit peak occurred at 20 min both in patients and in controls, whereas TSH and TSH-beta peaked at 60 min in patients and at 20 min in controls. One patient was given oral TRH (40 mg/day for 4 weeks). The beta/alpha ratio fell from 1.85 to 0.13. Interestingly, serum thyroid hormones, which did not increase after iv TRH and after the first doses of oral TRH, showed a definite increase. Sera from two patients were filtered on Sephadex G-100: in one of them TSH-beta eluted in the same position as labeled reference standard, whereas in the other one radioimmunoassayable TSH-beta eluted near the void volume. The above data indicate that in patients with idiopathic central hypothyroidism due to biologically inactive TSH there is an excess of circulating TSH-beta and suggest that TRH is implicated in the secretion of TSH of full biological potency.