We have tried to describe somatisation, not as a disease, but as a common and important human mechanism involving both doctor and patient. It is the single most common reason why psychiatric illness goes undetected in general medical settings, and it often occurs in conjunction with physical disease processes. The association with dysphoric affect has been recognised at least since George Cheyne 250 years ago, and the reason for this is that both anxiety and depression serve to amplify pains. However, it seems likely that somatisation can occur in the absence of dysphoria. Once it has been established, it is easy to see how it continues: it secures advantages from spouse, family and employers; and it tends to be encouraged by doctors--who differentially reward somatic symptoms. But why does it occur in the first place? We have argued that it seems to have three functions: First, it allows people who are unsympathetic to psychological illness, or who live in cultures where mental illness is stigmatised, to nonetheless occupy the sick-role while psychologically unwell. Second, it is blame-avoiding: instead of being responsible for the mayhem, one is cast in the role of the suffering victim. Finally, by reducing blame, it appears to save patients from being as depressed as they might otherwise have been.