Background The current study examines teacher ratings as a tool for id
entifying students at risk of developing psychosis. Follow-up and foll
ow-back studies have shown that teachers are capable of identifying in
dividuals who later develop serious mental illness. Method We examine
the long-term outcomes for individuals at genetic risk who were identi
fied as showing markedly deviant behaviour and those identified who di
d not show markedly deviant behaviour. Results Teachers were able to c
orrectly anticipate 35% of students who developed schizophrenia. Furth
ermore those identified as showing markedly deviant behaviour had poor
er clinical and psychiatric outcomes 10 and 25 years later than those
identified as not behaving with marked deviance.Their ratings also dif
ferentiated. within the group of people with schizophrenia. which indi
viduals would show evidence of poorer functioning 25 years later. Thes
e results were replicated in a group oi students not at genetic risk o
f schizophrenia.With in this low-risk group, teachers were able to pre
dict which students would develop psychotic disorders. Conclusions Tea
cher ratings were particularly useful in predicting clinical and psych
iatric outcomes 10 and 25 years later. The applicability of these find
ings in early intervention and treatment research is discussed.