Circadian rhythm of TSH secretion is characterized by a pronounced nocturnal increment that is not followed by the expected rise of circulating thyroid hormone levels. These findings suggest that the nocturnal TSH surge may be constituted by molecules with reduced bioactivity. We, therefore, investigated TSH bioactivity (measured as cAMP accumulation in FRTL-5 cells) and its carbohydrate structure (by Concanavalin A affinity chromatography) in different blood pools taken during the day and night from seven normal subjects and from one patient with mild (mPH) and five with severe primary hypothyroidism (sPH). Patients with sPH were also studied during low dose L-T4 treatment. Cosinor analysis showed a significant TSH circadian rhythm in the control group and in L-T4-treated sPH patients. The nocturnal TSH surge was not followed by any increase in free thyroid hormone levels. In normal subjects, the daytime ratio of TSH bioactivity to immunoreactivity (TSH B/I) was higher than the nocturnal one [1.4 +/- 0.6 (+/- SD) vs. 1.1 +/- 0.6; P < 0.02]. The same pattern was observed in the only mPH patient (1.0 +/- 0.2 vs. 0.7 +/- 0.1; P < 0.01), but not in the sPH patients (0.8 +/- 0.3 vs. 0.7 +/- 0.1; P = 0.3). L-T4 administration to sPH patients caused the daytime TSH B/I to increase and restored the day/night difference in the TSH B/I (1.0 +/- 0.3 vs. 0.8 +/- 0.3; P < 0.02). Concanavalin A chromatography showed that a higher percentage of less mature forms of TSH are secreted during the night. These data indicate that TSH molecules secreted during the night are less bioactive and differently glycosylated than those circulating in the same individual during the day, thus explaining why thyroid hormone levels do not rise after the nocturnal TSH surge. In sPH patients, the TSH circadian rhythm is restored during L-T4 administration, and day/night differences in TSH B/I similar to those recorded in normal subjects are observed.