5-min read
If you're asking does neurofeedback actually work, the honest answer is that the better a trial is at hiding who got the real treatment, the smaller the benefit looks. In the most carefully blinded results, participants showed no meaningful drop in ADHD symptoms. Older, less rigorous studies looked considerably more promising than that.
So does neurofeedback actually work for ADHD symptoms?
The most careful look at this pooled 38 randomized trials covering 2472 participants aged 5 to 40, all with a diagnosis of ADHD or a closely related condition (Westwood 2025). Most used the familiar version, where someone wears a cap of sensors on the scalp that reads electrical activity and gets feedback through a screen or a game. Thirty-three worked that way. Five used brain scanning instead.
The analysis then did something that sounds small and matters enormously. It separated out the ratings made by people who probably didn't know which participants had received the real treatment and which had received a comparison. Looking only at those ratings, total ADHD symptoms didn't improve in any meaningful way.
Two narrower results did show something. When the analysis looked only at trials using long established standard methods, a small improvement in total symptoms appeared. And among five different measures of thinking skills, one moved: the speed at which someone processes information. The other four didn't shift.
There's a catch worth holding onto. Most of the trials were judged to carry a high risk of bias, largely because they hadn't blinded people properly. That isn't a criticism of the analysis. It's a description of the material it had to work with.
One more thing about who was studied. Ages ran from 5 to 40, so this isn't a finding about children alone, and the analysis noted that too few trials in adults were available to say anything separate about them. For the wider ground on how attention gets studied, the focus and thinking section covers it.
What about the brain training apps and games?
Computerised brain training is a different product from neurofeedback, and it's been tested separately. One analysis pooled 36 trials with 2234 participants who had ADHD (Westwood 2023). Most of those trials recruited children aged 5 to 12. Eight recruited adults over 18, and only two recruited teenagers aged 13 to 18, which is worth knowing if the child you're thinking about is a teen.
The headline finding is that there's no evidence supporting computerised brain training as a treatment on its own for the core symptoms of ADHD. When the researchers looked at ratings from people who probably didn't know who got what, total symptoms didn't improve, and neither did hyperactivity and impulsiveness. Inattention showed a small improvement, and that improvement got noticeably bigger when it was rated in the same setting where the training happened. That pattern is a hint the benefit may not travel far beyond the room it was learned in.
Children did get better at the tasks the training drilled. Both kinds of working memory the researchers tested, remembering words and remembering where things are in space, improved in the shorter term. But those gains didn't carry over. Reading comprehension didn't move. Reading fluency didn't move. Arithmetic didn't move.
Two limits change how much weight this deserves. Most of the trials carried some or high risk of bias. And a lot of people dropped out: more than 20% of participants left in 17 of the 42 reports, which can make results look better than they are. Longer term effects rested on very few trials, so how long anything lasts is genuinely unresolved rather than settled.
Why the earlier picture looked more promising
Work published in 2011 reads far more encouragingly. A review examined 14 randomized studies of neurofeedback in young people with ADHD (Lofthouse 2011). Six had been published. Eight existed only as conference presentations that never reached a journal. Most ran between 1994 and 2010, in children aged 5 to 15, and the participants were usually boys and usually white, with the presentation of ADHD that combines inattention with hyperactivity.
Judged against a formal standard that professional bodies use for rating treatments, the review placed neurofeedback in the category of probably effective. Where improvement could be calculated, the average was what the reviewers classed as medium sized. Taken at face value, that sounds like a reason to try it.
Read the same review's own list of weaknesses and the picture shifts. Participants, the people rating them, and the trainers running the sessions often knew who was getting the real thing. There was no convincing sham version to compare against. Nobody was followed up after treatment ended, so nothing is known about whether gains held. Other conditions a child might have, and other treatments they might have been getting at the same time, weren't well accounted for. And for many of the studies, the numbers needed to calculate improvement were missing entirely.
That list matters more than the headline does. Those weaknesses are precisely what the later blinded analysis was built to strip out, and once they were stripped out, the group level benefit on ADHD symptoms went away. The two results aren't really in conflict. One shows what you see when raters know what happened. The other shows what's left when they don't.
What isn't known deserves saying plainly. Few of the pooled neurofeedback trials measured whether children were functioning better day to day, or how they were getting on at school (Westwood 2025). A question nobody has answered properly is not a question answered no. None of this tells you what to do for your child, and a decision about treatment belongs with the clinician who knows them.
Try this today: Write one sentence describing the single change you'd most want to see in your child, specific enough that their teacher would recognise it, and put today's date on it, so that any future claim about any program has something concrete to be measured against.
Sources
- Westwood S, Aggensteiner P, Kaiser A, et al.. Neurofeedback for Attention-Deficit/Hyperactivity Disorder. JAMA Psychiatry (2025). doi:10.1001/jamapsychiatry.2024.3702
- Westwood S, Parlatini V, Rubia K, et al.. Computerized cognitive training in attention-deficit/hyperactivity disorder (ADHD): a meta-analysis of randomized controlled trials with blinded and objective outcomes. Molecular Psychiatry (2023). doi:10.1038/s41380-023-02000-7
- Lofthouse N, Arnold L, Hersch S, et al.. A Review of Neurofeedback Treatment for Pediatric ADHD. Journal of Attention Disorders (2011). doi:10.1177/1087054711427530
[!note] Founder review required before publishing. Every claim above descends from a digest approved in KB Tab 3.