The administration of l-dopa suppresses prolactin (PRL) secretion in normal subjects and in patients with hyperprolactinemia, although it is not known whether this effect, which requires the conversion of dopa to dopamine, is mediated peripherally or through the central nervous system. To distinguish between these effects, 10 normal subjects (6 male, 4 female) and 8 patients with hyperprolactinemia associated with pituitary tumors were given l-dopa, 0.5 g alone, or 0.1 g after a 24-h pretreatment with carbidopa, 50 mg every 6 h, which produces peripheral dopa decarboxylase inhibition. Similar degrees of PRL suppression were observed in normal subjects (basal plasma PRL 13+/-2 ng/ml) after l-dopa alone (48+/-4%) and after l-dopa plus carbidopa (58+/-6%). In patients with pituitary tumors and elevated plasma PRL (73+/-14 ng/ml), l-dopa alone led to PRL suppression comparable with that in normal subjects (47+/-6%). However, l-dopa plus carbidopa resulted in only minimal suppression of plasma PRL (19+/-4%) which was significantly less than after l-dopa alone (P < 0.001). Urinary homovanillic acid excretion, which reflected peripheral dopa decarboxylation was similar in controls and tumor patients after l-dopa both alone and after carbidopa pretreatment. Comparable suppression of PRL levels in response to a dopamine infusion (4 mug/kg per min for 3 h) was observed in controls and tumor patients. The results indicate that although peripheral conversion of exogenous dopa to dopamine can suppress PRL secretion, in normals, the central nervous system conversion of dopa to dopamine in the presence of peripheral dopa decarboxylase inhibition is sufficient to account for its PRL-suppressive effects. In contrast, patients with tumors, while retaining peripheral dopaminergic inhibitory effects on PRL secretion, exhibit a marked reduction of central dopaminergic inhibition of PRL secretion.