The present study aimed to analyse the effect of risperidone on a priori defined core aggression items. Data were pooled from 163 boys (aged 5-12 years, with or without comorbid attention-deficit/hyperactivity disorder) with a DSM-IV diagnosis of either conduct disorder or oppositional defiant disorder who had participated in either of two identical, 6-week, randomized, double-blind, placebo-controlled trials. All received treatment with either placebo or oral risperidone solution (0.01-0.06 mg/kg/day). Subjects had below average intelligence [intelligence quotient (IQ) 36-84] and a score of > or =24 on the Conduct Problem subscale of the Nisonger Child Behaviour Rating Form (N-CBRF). An expert advisory panel selected six core aggression items from the N-CBRF, from which a total Aggression Score (AS, range 0-18) was constructed. Compared to those treated with placebo, risperidone-treated subjects experienced significantly greater mean decreases from baseline in the AS at each of weeks 1-6 (P<0.001). By study endpoint, aggression among risperidone-treated subjects had declined by 56.4% (mean baseline AS 10.1; mean endpoint AS 4.4), which was more than twice that of placebo-treated subjects (mean baseline AS 10.6; mean endpoint AS 8.3; 21.7% reduction). Risperidone was efficacious in reducing symptoms of aggression in boys of below average IQ with disruptive behaviour disorders.