The possible role of serotonin on prolactin (PRL) secretion was investigated by measuring PRL plasma levels before and after mechanical breast stimulation in puerperal women. Two serotonin antagonists, methysergide, an ergot-derivative, and pizotifen, a non-ergot compound, were used in the trial. After a control stimulation test on postpartum day 1, each mother received on day 2 at 7 a.m. 2 mg of methysergide (n = 7), 1 mg pizotifen (n = 6), or placebo (n = 6). At 9 and 9.30 a.m. after 30-min breast stimulation, blood was drawn for plasma PRL measurements. The test was repeated on postpartum day 5 after a 72 hour treatment with methysergide (7 cases) 1 mg q.i.d., pizotifen (6 cases) 0.5 mg q.i.d. and placebo (6 cases). Basal PRL was not modified by pizotifen, but treatment with methysergide induced a significant reduction (p < 0.001) on the 5th postpartum day. Pizotifen administration failed to significantly modify the transient hyperprolactinemia induced by suckling, but an inhibition of this effect by methysergide was conspicuous after acute administration (p < 0.05) and after 3-days treatment (p < 0.001). Since PRL suppression is common following the administration of ergot derivatives characterized as dopamine agonists, the inhibitory effect of methysergide on PRL secretion may be ascribed to a dopaminergic effect inherent in its ergot structure; in contrast, pizotifen, a non-ergot serotonin antagonist, was ineffective. However, since it is known that presently available compounds are not solely serotonin antagonists, the role of serotonin in the modulation of PRL secretion in man cannot be clarified until a pure antagonist drug is available.