I was pleased to read Tyrer's (Reference Tyrer2001) editorial arguing for the creation of a mixed category of anxiety and depression as a single diagnostic entity. Over two decades ago a paper of mine was published in theJournal reporting that attenders at a Maudsley Hospital out-patient clinic, when asked to check the symptoms experienced when they were depressed and when anxious, showed a correlation between the two mood states of 0.62. By contrast, ten experienced Maudsley psychiatrists, when asked to check the symptoms of a typical patient with a neurotic disorder, recorded a correlation of zero between anxiety and depression (Reference LeffLeff, 1978). While current diagnostic classifications perpetuate the problem, its origins would seem to lie in psychiatrists' training, with the promotion of textbook descriptions of mood states as ideal entities, bearing little relationship to the experiences of real-life patients.