Pyridoxine (vitamin B6) is metabolized in vivo to pyri-doxal-5′-phosphate, which serves as a cofactor for dopa decarboxylase. This enzyme converts dopa to dopamine, for which reason it has been thought that pyridoxine might function as a dopaminergic agent. The data concerning the effects of pyridoxine on pituitary hormones in normal subjects and in patients with hyperprolactine-mia have been conflicting. The present study was undertaken because of the lack of information concerning the effects of pyridoxine on gonadotropins and because of unexplained reports of the amelioration of galactorrhea by pyridoxine. The study was designed in an attempt to determine if acute or chronic pyridoxine administration could be shown to have any dopamine-like effect on the pituitary function in patients with amenorrhea and galactorrhea. The major emphasis, however, was to evaluate the effect of pyridoxine on prolactin and gonadotropins. Six patients (ages 22 to 37 years) with amenorrhea were studied, five of whom also had galactorrhea and elevated serum prolactin levels. Four provocative or suppressive tests were performed on alternate days after an overnight fast: TRH (500 mg) was injected as an intravenous bolus; L-dopa (500 mg) was given orally; chlorpromazine (25 mg) was injected intramuscularly (four patients); and a 300-mg intravenous bolus of pyridoxine was given. Blood samples were taken at intervals after each test. They were assayed for LH, FSH, and prolactin after L-dopa and chlorpromazine, and for these plus hGH and TSH after pyridoxine and TRH. The patients were then placed on 600 mg of pyridoxine per day orally (in three divided doses) and were readmitted 21 days later for a repeat of the above protocol on pyridoxine. They remained on 600 mg of pyridoxine per day orally for 3 to 4 months, with monthly clinical evaluations and basal serum prolactin, LH, and FSH measurements. The patients were similarly followed for 3 months after discontinuing pyridoxine. There were no significant responses of prolactin, LH, FSH, hGH, or TSH to pyridoxine under either basal or treatment conditions. TRH induced similar responses in plasma TSH before and during administration of pyridoxine. Before therapy, prolactin rose after TRH in four of the six patients. On pyridoxine, prolactin increased after TRH in two patients who had not responded initially. One of these women resumed menses while on pyridoxine. There was, however, no significant difference in mean prolactin reponse to TRH before or during pyridoxine therapy, whether comparison is made of maximum levels or increments of prolactin. The administration of 25 mg of chlorpromazine did not change the basal prolactin, LH, or FSH concentrations in any patient significantly. L-dopa caused a decrease in the basal prolactin concentration in all patients before pyridoxine treatment. The degree of suppression was unchanged after 3 weeks of pyridoxine therapy. Gonad-otropin levels did not change after L-dopa in either the pre- or post-therapy test. The L-dopa-induced rise in GH was unchanged before and after pyridoxine. Within individual tests, both before and after pyridoxine, prolactin increased after TRH and decreased after L-dopa, and hGH increased after L-dopa similarly. The means for each sampling time within tests were, however, significantly different (P < 0.05) for the TSH response to TRH. The mean basal prolactin, LH, and FSH (97.1 ± 14.8 ng/ml, 23.6 mlU/ml, and 18.8 ± 1.9 mlU/ml, respectively) during chronic pyridoxine treatment were not significantly different from pretreatment levels or from those 3 months after cessation of therapy. In addition, the mean basal prolactin of all tests after 21 days of pyridoxine (96.1 ± 10.3 ng/ml) was unchanged. Nevertheless, two of the six patients resumed regular menses during the chronic pyridoxine therapy. These patients had the lowest basal prolactin levels but could not be distinguished from the group on the basis of responses to the provocative tests nor on the basis of hormone levels after chronic therapy. One of these patients continued to have monthly menstrual periods during the 3 months after discontinuing pyridoxine, whereas the other had only two periods with light bleeding. Pretreatment levels of pyridoxal-5′-phosphate and hemolysate pyridoxal kinase were similar among the four patients in whom they were measured. After 3 to 4 months of pyridoxine treatment, the pyridoxal-5′-phosphate concentrations were also similar in the two patients who began menses (3 and 427 ng/1010 red blood cells; 89 and 156 ng/ml of plasma) and two of the patients whose menses did not begin (437 and 153 ng/1010 red blood cells; 126 and 67 ng/ml of plasma). The hemolysate pyridoxal kinase activity was likewise similar in the two pairs. No adverse effects were noted during pyridoxine therapy.