Background. The roofless have not been previously surveyed in detail l
ooking specifically at their psychiatric status. Method. All roofless
people in a discrete area were interviewed in their 'skippers' using a
semi-structured interview. Other roofless people helped in case findi
ng and diagnostic Validity was confirmed using a computerised case reg
ister. Results. Sixty-five people were interviewed. High rates of subs
tance abuse (n = 48, 75%), and very low Fates of psychotic illness (n
= 2, 3%) were found. High rates of serious physical morbidity, of impr
isonment and of being the victim of a crime were found. Conclusion. Th
e roofless have multiple severe medical. social, and psychiatric probl
ems. These all must be addressed when considering the provision of ser
vices.