Background. There has been an increase in the numbers of homeless young people in Britain. Little is known of the health and social welfare needs of this population.

Method. This case–control study compares a random sample of homeless people aged under 22 years recruited from consecutive attenders at two of London's largest facilities for homeless young people with a contemporaneous sample of domiciled young people recruited through general practice registration lists. The homeless and domiciled groups were compared on measures of childhood care, education and psychiatric disorder.

Results. One hundred and sixty-one homeless people (88% of those approached) and 107 domiciled subjects (60% of those approached) were interviewed. Sixty-nine per cent of homeless and a third of the domiciled subjects reported a childhood lacking in affection, with indifferent and often violent carers. Psychiatric disorder was identified in 62% of homeless respondents and a quarter of the domiciled population. A fifth of homeless and 5 domiciled respondents had attempted suicide in the previous year. Multivariate analysis suggest that childhood adversity, low educational attainment and the prior presence of psychiatric disorder all independently increase the likelihood of homelessness in a youthful population.

Conclusions. The evidence presented in this paper supports the hypotheses that characterize the young homeless population as experiencing higher rates of childhood adversity and psychiatric disorder than their domiciled contemporaries. A tentative model is suggested whereby childhood experiences, educational attainment and the prior presence of psychiatric disorder all independently increase the likelihood of homelessness in a youthful population.