We administered the serotonin-1a agonist buspirone (0.4 mg/kg orally) as a neuroendocrine challenge agent to a group of male patients with DSM-III-R major depressive disorder (MDD) (n = 13) and a group of male healthy controls (n = 10). The primary hypothesis of the study was that the prolactin response to buspirone would be blunted in the depressed patients. The prolactin response was significantly lower in depressed patients than in controls. There was no significant relationship between placebo corrected-peak prolactin level and severity of depression or suicidality. There was a nonsignificant trend for the melancholic (n = 5) depressed patients to have a lower placebo corrected-peak prolactin level than nonmelancholic depressed patients (n = 8). Our findings support a role for the serotonin-1a receptor in the etiology of MDD, specifically at the postsynaptic site.