Introduction Classroom-based active breaks are a feasible and effective way to reduce and break up sitting time, and to potentially benefit physical health in school children. However, the effect of active breaks on children’s cognitive functions and brain activity remains unclear. Objective We investigated the impact of an active break intervention on typically developing children’s cognitive functions and brain activity, sitting/standing/stepping, on-task behaviour, and enjoyment. Methods Up to 141 children, aged between 6 and 8 years (46% girls), were included, although about half of them completed two of the assessments (n = 77, working memory; n = 67, dorsolateral prefrontal cortex haemodynamic response). Classrooms from two consenting schools were randomly allocated to a six-week simple or cognitively engaging active break intervention. Classrooms from another school acted as a control group. The main analyses used linear mixed models, clustered at the class level and adjusted for sex and age, to investigate the effects of the interventions on response inhibition, lapses of attention, working memory, event-related brain haemodynamic response (dorsolateral prefrontal cortex). The mediating effects of sitting/standing/stepping on cognition/brain activity were also explored. To test intervention fidelity, we investigated differences by group on the change values in children’s sitting, standing, and moving patterns during class/school time using linear mixed models. Generalized linear mixed models clustered at the individual level were used to examine ontask behaviour data. For the intervention groups only, we also assessed children’s perceived enjoyment, physical exertion and mental exertion related to the active breaks and compared the results using independent t-tests. Results There was a significantly greater positive change in the proportion of deoxygenated haemoglobin in the left dorsolateral prefrontal cortex of children assigned to cognitively engaging active breaks compared to the control group (B = 1.53 × 10-07, 95% CI [0.17 × 10-07, 2.90 × 10-07]), which under the same cognitive performance is suggestive of improved neural efficiency. Mixed models showed no significant effects on response inhibition, lapses of attention, working memory. The mediation analysis revealed that the active breaks positively affected response inhibition via a change in sitting and standing time. The sitting, standing, and moving patterns and on-task behaviour were positively affected by the active breaks at end of trial, but not at mid-trial. Children in both intervention groups showed similarly high levels of enjoyment of active breaks. Conclusion Cognitively engaging active breaks may improve brain efficiency in the dorsolateral prefrontal cortex, the neural substrate of executive functions, as well as response inhibition, via effects partially mediated by the change in sitting/stepping time. Active breaks can effectively reduce sitting and increase standing/stepping and improve on-task behaviour, but the regular implementation of these activities might require time for teachers to become familiar with. Further research is needed to confirm what type of active break best facilitates cognition.