BACKGROUND: The randomised controlled trial (RCT) design is increasingly common among studies seeking good-quality evidence to advance educational neuroscience, but conducting RCTs in schools is challenging. Fit to Study, one of six such trials funded by the Education Endowment Foundation and Wellcome Trust, tested an intervention to increase vigorous physical activity during PE lessons on maths attainment among pupils aged 12-13. This review of designing and conducting an RCT in 104 schools is intended as a resource on which researchers might draw for future studies. METHOD: We consider intervention design and delivery; recruitment, retention, trial management, data collection and analysis including ethical considerations and working with evaluators. RESULTS: Teacher training, intervention delivery and data collection during large-scale RCTs require a flexible approach appropriate to educational settings, which in turn entails planning and resources. CONCLUSION: Simple interventions, with few outcome measures and minimal missing data, are preferable to more complex designs.