It is established that the mildly learning disabled population has a three fold elevated risk for schizophrenia. On the basis of findings from previous neuroimaging studies it has been proposed that in some individuals learning disability is a prepsychotic manifestation of schizophrenia. On this background, a cohort was selected from a nonpsychotic adolescent population in special education by employing tools previously shown to identify those at elevated risk of schizophrenia from within a high risk population. The risk of developing schizophrenia within this selected cohort was expected to be substantially greater than that for the learning disabled population as a whole. This population was then assessed by clinical interview, neuropsychological assessment and MRI scanning. Region of interest methodology was employed to ascertain amygdala volume in both the high risk and a matched control group. Two primary areas of interest were addressed; comparison of amygdala volume between the two groups and investigation of the relationship between symptomatology and amygdala volume within the high risk population. While no significant difference was found between amygdala volume in the high risk and control groups, a significant negative correlation was seen between left amygdala volume and weight of negative symptoms within the high risk group (p=0.009). This suggests that within this population reduced amygdala volume may be significant in the aetiology of negative-type symptoms and these symptoms may be present prior to clinical illness.