Adrenocorticotropic hormone (ACTH), cortisol, and prolactin responses following maximal and submaximal (40 min at 80% maximal O2 consumption) running were studied in eumenorrheic (ER; n = 8, 29.0 +/- 1.5 yr) and amenorrheic (AR; n = 8, 24.5 +/- 2.0 yr) runners. ER were studied in the early follicular and midluteal phases of the menstrual cycle. Physical, training, and gynecological characteristics were similar, and cardiorespiratory and metabolic responses to the exercises were indistinguishable in the groups. ACTH, cortisol, and prolactin data from the follicular luteal phases in ER were combined for comparison to AR, because no differences were noted between the menstrual phases at rest. Similar preexercise ACTH levels and responses following exercise occurred in both groups, but preexercise cortisol levels were elevated (ER = 293.1 +/- 46.3, AR = 479.6 +/- 42.4 nmol/l) and cortisol responses blunted in AR. Adrenal sensitivity was blunted in AR compared with ER after submaximal (ER = 121.9 +/- 17.4, AR = 51.7 +/- 13.6) and maximal exercise (ER = 27.9 +/- 9.2, AR = 12.1 +/- 3.8). Preexercise prolactin levels were reduced (ER = 16.4 +/- 2.7, AR = 10 +/- 2.3 micrograms/l), and prolactin responses to maximal exercises were blunted in AR, despite high lactate levels (11.4 +/- 0.4 mmol/l). We conclude that 1) control for menstrual phase in ER is important in studies of prolactin responses following exercise but not in studies of ACTH and cortisol responses following exercise, 2) cortisol responses following submaximal and maximal exercise in AR are blunted at the adrenal level, 3) prolactin responses following submaximal and maximal exercise are also blunted in AR, and 4) prolactin responses following exercise may be mediated by adrenal activation.