We report here on a study conducted to measure the validity of an Amharic version of the Reporting Questionnaire for Children (RQC), that was used in a survey of childhood behavioural disorders in a predominantly rural district in western Ethiopia. Mothers of 196 children aged 5-15 years, who were initially interviewed by the RQC were re-interviewed by a psychiatrist who was unaware of the RQC status of these children. The re-interview was conducted using a DSM IV checklist. The study showed that a cut-off point of one or more positive responses to any of the 10 questions on the RQC maximized sensitivity (87.5%) and specificity (65%). The discriminatory power of each item was also computed, and the item dealing with wetting/soiling oneself was found to have the highest ability to identify cases from non-cases. The item on abnormal speech was found to have the least discriminating power.