Depression may be associated with a hypofunction of central serotonergic systems. The prolactin response to fenfluramine, an indicator of serotonergic activity, has been reported to be blunted in depressed patients compared to controls. It has also been suggested that blunting is more likely in melancholic depression. Baseline cortisol, prolactin and tryptophan availability have also been suggested to affect this response. Forty-eight men and 61 women with a major depressive episode, and who were drug free, and 20 healthy control men underwent clinical evaluation and fenfluramine challenge with dl-fenfluramine 1 mg/kg. When baseline variables were covaried, there was no difference in prolactin response to fenfluramine between males with depression and age-matched controls. Amongst all the depressed patients, body mass index showed a significant association with prolactin response to fenfluramine. There was an interaction between baseline cortisol and DSM-III-R melancholic subtype of depression whereby non-melancholic patients appeared more likely to increase prolactin response to fenfluramine in response to higher cortisol levels. Prolactin response to fenfluramine was not blunted in major depression and there was no difference between melancholic and non-melancholic depression. However, the relationship between prolactin response to fenfluramine and baseline cortisol levels appeared to differ between these two subtypes of depression.