Twenty-seven women with secondary amenorrhea of greater than six months duration were subjected to multiple testing of hypothalamo-pituitary function. They were divided into normo-prolactinemic (Group 1 mean serum prolactin (PRL) 9.8 ng/ml; range 6.8 to 13.0 ng/ml; n=9) and hyperprolactinemic (Group 2 mean 37.5 ng/ml; range 19.2 to 93.7 ng/ml; n=18) groups on the basis of 4 weekly baseline determinations. Group 2 had significantly (P less than .05) lower serum LH and urinary pregnanediol levels than did Group 1; there was no statistical difference between the groups in serum FSH, T4, T3 or urinary estrogen measurements. Two women in Group 2 were found to have a pituitary chromophobe adenoma. Group 2 women showed no significant rises in serum PRL following stimulation tests with thyrotropin releasing hormone (TRH, 200 microng iv) and metoclopramide (10 mg orally), which caused significant responses in Group 1. The TSH response to TRH was, however, preserved in Group 2, while it was subnormal in Group 1 subjects. Both groups showed similar FSH and LH responses to luteinizing hormone-releasing hormone (LHRH, 25 microng iv). No significant suppression of serum PRL was seen in Group 2 patients given L-Dopa (500 mg orally),, which produced a significant response (P less than 0.05) in Group 1 subjects, while all patient showed marked reduction in serum PRL values following 2-bromo-alpha-ergocryptine (CB-154, 2.5 mg orally). When compared with other Group 2 members, the 2 cases with proven pituitary adenomata gave similar responses to the stimulation-inhibition tests and were not clearly distinguished on this basis. We conclude: 1. The pattern of PRL responses to dynamic tests, although of pathophysiological interest an autonomous pituitary lesions in patients with hyperprolactinemic secondary amenorrhea. 2. Such dynamic tests, although a pathophysiological interest, provide no clinical information additional to that provided by the mean basal serum PRL value. 3. In clinical practice, such dynamic tests should be confined to patients with mean serum PRL levels at around the upper limit of the normal range.