Population-based epidemiological surveys of deliberate self harm (DSH) exclude homeless people and do not represent them in their conclusions. However, DSH in the homeless is commonly seen in inner city hospitals and the homeless are probably at higher risk of eventual suicide. We examined the demographic characteristics of all patients who presented to an inner London hospital after an episode of DSH over a period of three years and found that 15% of patients were of no fixed abode. Alternative methods of studying DSH need to be explored if suicide is to be prevented in this group of people.