Emotional and behavioural difficulties (EBD) are common in children, and forecasting their prevalence in schools is of interest to both academic researchers and local authorities. Percentage of pupils eligible for free school meals is one measure often used for this purpose. The article presents the first independent validation of a simple numerical score, the Ford score, which predicts the level of EBD within schools using routinely collected school variables on eligibility for free school meals, special educational needs, unauthorised absences and permanent exclusions. The research tested the predictive utility of the Ford score in the English sub-sample of the British Child and Adolescent Mental Health Survey, 2004. This gave a sample of 6,379 children aged 5-16 in 3,117 schools, with mental health being assessed through clinician-assigned psychiatric diagnoses and through brief questionnaires filled in by parents, teachers and young people. The Ford score was found to be highly predictive of the level of EBD by all four of these measures, with a linear effect across the whole range. The estimated prevalence of mental disorder rose from about 5 per cent in schools with low Ford scores to around 14 per cent in high-scoring schools. The Ford score was more highly correlated with all measures of EBD than percent-age of free school meals alone. The Ford score's predictive utility did not seem to be further enhanced, however, by adding in an additional measure of local area deprivation. It was concluded that the Ford score represents a simple and up-to-date school EBD predictor which is more accurate than percentage free school meals alone and which is not notably improved by taking account of a schools' area deprivation. (Contains 2 tables and 4 figures.)