The distinction between non-psychotic repressive illness and anxiety states is blurred. Large scale trials in the neuroses indicate that benzodiazepines are ineffective in depression, and transiently and partially effective in anxiety. In contrast, tricyclic antidepressants are effective in both. All effective antidepressants decrease 5HT2 receptors number and this may mediate antidepressant efficacy. Our studies indicate that reduction of 5HT2 relative to 5HT1 neurotransmission would reverse the neuroendocrine abnormalities we have described in depression. Reduced 5HT2 neurotransmission may also be a mechanism of anxiolytic action in view of 5HT theories of punishment. There is clinical evidence for anxiolytic and antidepressant action of selective 5HT2 antagonists.