Children with cerebral palsy (CP) experience deficits in nonverbal reasoning. The SMART online cognitive intervention has been associated with gains in IQ and nonverbal IQ in previous studies in typically developing school-aged children and children experiencing learning difficulties. To assess the efficacy of an online cognitive intervention in school-aged children with CP. 21 children with CP (male n = 17; 76.2%), mean age 9 y 8 m, SD 1 y 1 month (range 8 y 3 m to 12 y 6 m) were randomised into the intervention group (n = 9) or a waitlist control group. A mixed-methods approach with an explanatory sequential design was used, with a randomised controlled trial followed by qualitative interviews. Participants were assessed on measures of intelligence, academic ability, attention and executive functioning, and social-emotional functioning at baseline, then after completing the training, or the waitlist period. Analyses included ANCOVAs and paired samples t tests. Semi-structured interviews explored participants’ experiences with the training. Training completion was low with a mean of 16.9 modules completed out of 55 available. No significant effect of training was found for the primary outcome of intelligence, or for any secondary outcomes. Participants reported barriers and facilitators for accessing the program. Cognitive training programs addressing relational framing ability may require significant modifications before they can be effectively tested with children with CP. •RCT of a novel online cognitive intervention in children with cerebral palsy.•Intervention grounded in relational frame theory to train relational abilities.•Qualitative data on acceptability and accessibility of online intervention.•Results highlight importance of targeted, individualised intervention.•Concrete and visual aids may assist children to engage with training.