Seven studies were included. Four randomised controlled trials were analysed, with a total of 191 participants. For quality, one trial was rated as type one, and the other three were rated as type two.
There were mixed effects for the benefit of CWMT for children with ADHD.
Compared with placebo (a low-level working memory version of CWMT), CWMT at three months had a moderate effect on trained and untrained cognitive tasks, and on parent-reported symptoms (one trial). In another trial using the same placebo, statistically significant effects in favour of CWMT were reported only for trained cognitive tasks and behavioural observations.
Compared with a waiting-list control, CWMT had a large effect on parent-reported symptoms, with no statistically significant differences for any other domains (one trial).
Compared with an active comparison of Academy of Math, CWMT had a small effect on trained cognitive tasks only (one trial).
The other three studies (one randomised controlled trial and two before-and-after studies) were described separately. They did not have ADHD symptom outcomes, and did not include participants with a formal diagnosis of ADHD.