The impact of chronic high volume athletic training on thyroid hormone economy has not been defined. We investigated the status of the hypothalamic-pituitary-thyroid axis (H-P-T) in women athletes with regular menstrual cycles (CA) and with amenorrhea (AA). Their data were compared with each other and with those derived from cyclic sedentary women (CS) matched for a variety of confounding factors including the intensity of exercise, caloric intake, and body weight. Alterations of the H-P-T axis were observed in women athletes compared to CS. While serum levels of T4, T3, free T4, free T3 and rT3 were substantially reduced (P less than 0.01) in AA, only serum T4 levels were significantly decreased in CA. Further, remarkable differences were found between CA and AA in that serum levels of free T4 (P less than 0.01), free T3 (P less than 0.01), and rT3 (P less than 0.05) were significantly lower in AA than in CA. Thyroid binding globulin and sex-hormone binding globulin concentrations were within their normal ranges for all groups of subjects. Both 24-h mean TSH levels and the circadian rhythm of TSH secretion were also comparable. However, the TSH response to TRH stimulation was blunted (P less than 0.01) in AA when compared to CA, but not to CS. Whereas the underlying mechanism(s) to account for the "global" reduction of circulating thyroid hormone in the face of normal TSH levels in AA is presently unknown, these observations provide information of clinical significance: 1) chronic high volume athletic training in women athletes with menstrual cyclicity is accompanied by an isolated T4 reduction; 2) an impaired H-P-T axis occurs selectively in athletic women in whom chronic high volume athletic training is associated with compromised hypothalamic-pituitary-ovarian function and amenorrhea.