Do brain training games work? The short answer
Do brain training games work? You get better at the games you practise, but the best review of the evidence found little proof that this carries over to everyday memory, attention or thinking. One long trial found a lower dementia risk for one specific type of speed training in older adults, which is promising but narrow. Treat these apps as entertainment, not medicine.
That answer is less satisfying than the adverts, so it is worth looking at how we got there. This piece walks through the 2016 review that most researchers point to, the regulator action that followed, the single trial that complicates the story, and a simple way to judge any brain training claim you meet.
A note on limits before we start: everything here is about computerised brain training in general, not a ranking of specific apps. We have not tested any product, and we do not claim any of them will improve your memory.
What the 2016 review actually found
The most cited assessment is a review in Psychological Science in the Public Interest, published in 2016 and led by University of Illinois researchers Daniel Simons and Elizabeth Stine-Morrow, with colleagues from Florida State, Michigan State, Union College and Cambridge's Medical Research Council. According to ScienceDaily's coverage, the team examined 132 journal articles cited by brain-training companies, plus further studies found on company websites.
Their conclusion, in the press summary, was blunt: 'There is little evidence to date of real-world benefits from brain training.' The review separated three things. There was extensive evidence that people improve on the tasks they train. There was less evidence that they improve on closely related tasks. And there was little evidence that training improves distantly related tasks or everyday cognitive performance.
That ladder is the most useful idea in the whole debate. A claim that you will get faster at a puzzle is easy to meet. A claim that you will remember your shopping, or cope better at work, needs evidence that the improvement transfers, and that is where most of the studies fell short.
“If you practice remembering playing cards, you'll get really good at remembering playing cards. But does that help you remember which medications to take, and when?”
Why the studies fell short: weak controls and narrow tests
The review did not just say the results were weak. It listed why the evidence was hard to trust. ScienceDaily's summary names small sample sizes, selective reporting of significant results, and the same study results appearing across several papers without acknowledgement. It also flagged inadequate control groups: some studies used passive controls, some used none, and some used mismatched controls.
The control group matters more than it sounds. If one group plays a brain training game and another does nothing, any difference could come from expectation, attention or simple novelty. A fairer test gives the comparison group an equally engaging activity. Many of the studies the review examined did not do that.
The other problem was what got measured. Many studies tested abstract lab tasks rather than real-world performance. Daniel Simons put the problem plainly: 'If you practice remembering playing cards, you'll get really good at remembering playing cards. But does that help you remember which medications to take, and when?'
The Lumosity settlement: when the claims got ahead of the science
The regulator action fits the same story. On 5 January 2016 the US Federal Trade Commission announced that Lumosity would pay $2 million to settle charges of deceptive advertising. The FTC alleged the company made unfounded claims that its games could improve performance at work and school, reduce cognitive decline, and protect against dementia and Alzheimer's disease.
The settlement also included a $50 million judgment that was suspended because of the company's financial hardship. Lumosity had to notify subscribers who signed up between 2009 and 2014 about the action and give them ways to cancel, and future health claims must be backed by competent and reliable scientific evidence. Two co-founders, Kunal Sarkar and Michael Scanlon, were named as individual defendants.
Jessica Rich, then director of the FTC's Bureau of Consumer Protection, said Lumosity 'preyed on consumers' fears about age-related cognitive decline.' The FTC's case was about advertising claims, not a ruling on whether every brain game is useless. But it is a useful reminder that marketing about memory and dementia ran well ahead of the evidence.
The one trial with a dementia signal: ACTIVE
There is a genuine exception to the pessimistic picture. The Advanced Cognitive Training for Independent and Vital Elderly study, known as ACTIVE, enrolled 2,802 healthy older adults averaging age 74 at six US sites. Participants were randomised to memory strategy training, reasoning strategy training, computerised speed-of-processing training, or a no-training control. The speed group did 10 initial sessions over six weeks, and some got four extra booster sessions.
A paper led by Jerri Edwards, published in 2017 in Alzheimer's & Dementia: Translational Research & Clinical Interventions, reported that after 10 years the speed-training group had a 29% lower risk of dementia than the control group (hazard ratio 0.71, 95% confidence interval 0.50 to 0.998). Memory and reasoning training did not produce a significant effect. Among people who completed 15 or more sessions, dementia rates were 5.9% for speed training, 9.7% for memory, 10.1% for reasoning and 10.8% for controls, per ScienceDaily.
Edwards noted a dose effect: 'When we examined the dose-response, we found that those who trained more received more protective benefit.' That is encouraging. It is also a long way from proof that a phone game will protect you.
What the ACTIVE result does not show
Read the numbers carefully. The confidence interval runs from 0.50 to 0.998, so the result sits very close to the line where an effect would no longer be statistically significant. Only one of the three training types showed a benefit. Researchers also stressed that the training is preventative for lowering dementia risk, not a treatment for people who already have dementia.
The trial tested one specific computerised speed-training programme in older adults who were healthy at the start, under a structured schedule with trained staff. It does not show that other brain games, other ages or casual play deliver the same result. We have also not opened a replication of this finding for this article, so treat it as one trial, not settled science.
The honest way to hold the two pieces of evidence together is this. The broad review says most brain training does not transfer to daily life. The ACTIVE trial says a specific kind of training may be worth further study in older adults. Neither supports the idea that any brain game makes you smarter.
How to judge a brain training claim
Ask what was measured. If the evidence is that people got better at the game, that is the easy result. If the claim is better memory, focus or protection from dementia, look for a study that measured those outcomes in daily life, not just a different puzzle.
Ask who the control group was and how many people took part. A small study with no active control is weak evidence, which is exactly the pattern the 2016 review criticised. Then check who funded and published it. Studies cited by the companies selling the product need more scrutiny, not less.
Finally, ask what you would lose if the claim is wrong. If a game is cheap, fun and you enjoy it, the downside is small. If you are paying a subscription because you fear memory loss, that is the situation the FTC action was about, and it is worth asking a doctor rather than an app.
Our take
Brain training games work at making you better at brain training games. The evidence that they make you better at life is thin, and the best review of it said so in plain words. The ACTIVE trial is the strongest counterpoint, but it is one trial, one programme and a borderline result.
If you enjoy puzzles, play them for the fun. Do not pay a premium for a promise about memory or dementia, and be wary of any app that makes one. We cover puzzle and brain games on their merits in our brain section, and we will not call any of them a health product.
Frequently asked
Do brain training games work?
They make you better at the games you practise, but a 2016 review in Psychological Science in the Public Interest found little evidence of real-world benefit after examining 132 journal articles cited by brain-training companies. Improvement on closely related tasks had less support, and improvement on everyday cognitive performance had little. One trial, ACTIVE, found a lower dementia risk after speed training in older adults, but that result is narrow and borderline.
Can brain training games prevent dementia?
There is no general proof. The ACTIVE trial enrolled 2,802 healthy adults averaging age 74 and reported a 29% lower dementia risk after 10 years for computerised speed training (hazard ratio 0.71, interval 0.50 to 0.998). Memory and reasoning training did not show a significant effect. Researchers described the training as prevention, not a treatment for existing dementia, and the result applies to that specific programme, not to games in general.
Why did the FTC fine Lumosity?
On 5 January 2016 the FTC announced Lumosity would pay $2 million to settle deceptive advertising charges. It alleged the company made unfounded claims that its games could improve work and school performance, reduce cognitive decline and protect against dementia and Alzheimer's. A $50 million judgment was suspended because of financial hardship. The company also had to notify subscribers from 2009 to 2014 and back future health claims with competent scientific evidence.
What is the problem with brain training studies?
The 2016 review, via ScienceDaily, listed small sample sizes, selective reporting of significant results, and the same results appearing across multiple papers without acknowledgement. It also flagged inadequate control groups, including passive, absent or mismatched controls, and a focus on abstract lab tasks rather than real-world performance. Together those issues make it hard to tell whether any improvement comes from the training itself or from expectation and practice.
Do brain games improve memory?
The evidence says they improve memory for the thing you practise, not memory in general. Daniel Simons gave the example that practising remembering playing cards makes you good at remembering playing cards, but does not clearly help you remember which medications to take and when. The 2016 review found little evidence that training improves distantly related tasks or everyday cognitive performance, so do not expect a puzzle app to fix daily forgetfulness.
Is it worth paying for a brain training app?
Only if you enjoy it as entertainment. The 2016 review found little real-world benefit, and the FTC's Lumosity case centred on health claims that were not backed by scientific evidence. If you are paying because you worry about memory or dementia, the evidence does not support that purchase, and a doctor is a better first step. A cheap puzzle game you like is a different decision from a subscription sold on fear.