Female Sprague-Dawley rats were ovariectomized, given estrogen, and blood samples were obtained via an atrial catheter in the afternoon during the prolactin (PRL) surge. Restraint stress applied at 16.00 h and continued for 3 h resulted in marked decrease in plasma prolactin (PRL) and an increase in plasma corticosterone (B). The neural mechanism(s) involved in the plasma PRL decrease to restraint stress in the afternoon were examined using neural agonists and antagonists. The administration of pimozide, a dopamine antagonist, increased plasma PRL and completely prevented the restraint-induced decrease in PRL. This result suggested that an increase in dopamine secretion mediated the stress-induced decrease of PRL in the afternoon. In unrestrained animals, the intravenous administration of atropine (a muscarinic cholinergic antagonist), arecoline (a muscarinic cholinergic agonist), propranolol (a β-adrenergic antagonist) and morphine (a β-endorphin agonist) at 16.00 h decreased plasma PRL from that of vehicle-injected animals. Bicuculline (a GABAergic antagonist) had no effect while phentolamine (an α-adrenergic antagonist) and phenoxybenzamine (an α-adrenergic antagonist) initially increased and then decreased plasma PRL. Naloxone (a β-endorphin antagonist) initially decreased and then increased plasma PRL in unrestrained animals. In restrained animals, the intravenous administration of atropine and naloxone had no effect on the decrease in plasma PRL. Bicuculline and propranolol decreased plasma PRL below that observed for restrained animals alone, while phentolamine and morphine slightly retarded the course of the decrease. Arecoline did not alter the PRL decrease to restraint in the early sample periods but was followed by a rebound increase at later times. Phenoxybenzamine initially elevated the PRL level above that observed at 16.00 h and then the PRL level gradually decreased. Plasma B was similar to control values in both unrestrained and restrained animals when they were administered atropine, arecoline, bicuculline, propranolol, naloxone and morphine. The administration of phentolamine and phenoxybenzamine significantly elevated plasma B in both unrestrained and restrained animals when compared to control values. The data indicated that neural mechanism(s) involved in the stress-induced decrease of PRL in the afternoon differed from that of the stress-induced increase of PRL observed in the morning of ovariectomized, estrogen-treated