This paper presents the research protocol for a pragmatic randomised controlled trial of the PAX ‘Good Behaviour Game’ in disadvantaged schools. The PAX ‘Good Behaviour Game’ (GBG) is an evidence-based universal prevention programme that blends its themed activities and materials with normal class instruction. The main outcome of the programme is to teach the children the mental ability to self regulate and in doing so, promoting positive behaviour in the classroom. The protocol outlines a research design that will assess whether the GBG programme over a 12-week period can improve a number of specific outcomes in primary 3 children, in a sample of schools identified as disadvantaged. The outcome measures will include both child and teacher reported behaviours, assessing several primary and secondary outcomes related to self-regulation, cooperative learning, self-esteem and pro-social behaviour, including a process evaluation to measure the fidelity, delivery and acceptability of programme and control schools’ activities.