The inhibitory effect of nomifensine ( Nom ; 200 mg orally) on prolactin (PRL) secretion was studied in 15 subjects with puerperal hyperprolactinemia and in 59 pathologic hyperprolactinemic women. The latter were grouped as follows: i) patients with surgically proven PRL secretory pituitary adenomas (proven tumors; 27 cases); ii) patients presenting radiological signs of sella indicative of a pituitary tumor (presumptive tumors; 10 cases); iii) subjects with non-drug induced hyperprolactinemia (hyperprolactinemia of uncertain etiology; 22 cases). A mean PRL fall of 30% or more of baseline hormone levels in samples collected within the 120-240 min post-treatment interval was adopted to define responsiveness to Nom . In 24 out of 27 subjects with proven tumors and in 9 out of 10 subjects with presumptive tumors Nom did not induce significant variations in PRL secretion. In only 11% of the patients with surgery-confirmed or highly suspected tumors a hormone decrease greater than 30% was observed. In addition, 13 subjects with hyperprolactinemia of uncertain etiology did not respond to Nom administration. In 5 of these, additional data suggesting the existence of an adenoma were collected. Finally, 3 out of 9 Nom -responder patients presented either a polycystic ovary syndrome or transitory hyperprolactinemia. The finding that hyperprolactinemic women, who did not show clinical or radiological signs of a tumor and patients with highly presumptive or proven pituitary tumors may present comparable responses to Nom , suggests that this pattern may be indicative of an early manifestation of a PRL-secreting adenoma which has yet to evolve. The follow-up of Nom -non-responder hyperprolactinemic subjects who did not show clinical signs of harboring a tumor, is therefore advisable.