BACKGROUND: We aim to evaluate the effectiveness of the Good School Toolkit, developed by Raising Voices, in preventing violence against children attending school and in improving child mental health and educational outcomes. METHODS/DESIGN: We are conducting a two-arm cluster randomised controlled trial with parallel assignment in Luwero District, Uganda. We will also conduct a qualitative study, a process evaluation and an economic evaluation. A total of 42 schools, representative of Luwero District, Uganda, were allocated to receive the Toolkit plus implementation support, or were allocated to a wait-list control condition. Our main analysis will involve a cross-sectional comparison of the prevalence of past-week violence from school staff as reported by children in intervention and control primary schools at follow-up.At least 60 children per school and all school staff members will be interviewed at follow-up. Data collection involves a combination of mobile phone-based, interviewer-completed questionnaires and paper-and-pen educational tests. Survey instruments include the ISPCAN Child Abuse Screening Tools to assess experiences of violence; the Strengths and Difficulties Questionnaire to measure symptoms of common childhood mental disorders; and word recognition, reading comprehension, spelling, arithmetic and sustained attention tests adapted from an intervention trial in Kenya. DISCUSSION: To our knowledge, this is the first study to rigorously investigate the effects of any intervention to prevent violence from school staff to children in primary school in a low-income setting. We hope the results will be informative across the African region and in other settings. TRIAL REGISTRATION: clinicaltrials.gov NCT01678846.