Does Brain Training Work? What the Research Actually Says

By Dawid Siedlarz · · 7 min read

Play Schulte Table free →

Short answer: brain training makes you better at the games you train and at very similar tasks. The large studies and reviews (the 2010 Nature study with 11,430 participants, the ACTIVE trial with 2,832 older adults, and the 2016 review by Simons and colleagues) found little evidence that it raises general intelligence, improves everyday memory or prevents dementia. In 2016 the maker of Lumosity paid 2 million dollars to settle US charges that its ads claimed more. Exercise, sleep, social contact and learning a demanding new skill have better evidence. Brain games are still worth playing: for enjoyment, focus and a measurable sense of progress.

"Train your brain in ten minutes a day" is one of the most successful slogans in the app market. It promises something everyone wants: a sharper memory and a mind that ages more slowly, delivered by a game. The question is whether the science behind it holds up. Researchers have tested brain training for more than twenty years, with tens of thousands of participants, and the answer is more specific than either the ads or the sceptics suggest.

What "brain training" means

Brain training (also called cognitive training) means practising short, repeatable tasks that exercise one mental ability: remembering a sequence, finding targets quickly, keeping several things in mind, switching between rules or doing arithmetic under time pressure. Commercial apps package these tasks as games with scores and levels. Classic puzzles like Sudoku and crosswords are often counted too.

The key scientific idea is transfer:

  • Near transfer: getting better at tasks very similar to the one you practised. After a week of Schulte tables you will be faster at a new Schulte table, and probably at other "find the targets" tasks.
  • Far transfer: getting better at something different that was not practised, such as reasoning, school results, remembering where you left your keys, or general intelligence.

Practice always produces some near transfer; that is simply learning. The big claim of brain training is far transfer, and that is what the research has tested.

What the major studies found

StudyWho and whatResult
ACTIVE trial (USA, started 1998)2,832 adults aged 65 and over; about ten sessions of memory, reasoning or speed training, plus boostersEach group improved in the ability it trained, and the gains were still measurable years later. Little improvement in the abilities that were not trained.
Owen et al., Nature (2010)11,430 adults who trained online for six weeks, recruited through a BBC programmeClear improvement on the trained tasks; no transfer to untrained tests of reasoning and memory, even closely related ones.
Jaeggi et al., PNAS (2008)Small study of the "dual n-back" working memory taskReported gains in fluid intelligence. Later meta-analyses (for example Melby-Lervåg and Hulme, 2013) found no convincing far transfer.
Simons et al., Psychological Science in the Public Interest (2016)Review of the studies cited by brain-training companiesGood evidence for improvement on trained tasks, less for closely related tasks, little for distant tasks or everyday cognitive performance. Many studies had small samples or weak control groups.
FINGER trial (Finland, 2015)1,260 people aged 60 to 77 at risk of dementia; two years of diet, exercise, cognitive training and blood-pressure care togetherBetter scores on cognitive tests than the control group. Cognitive training was one part of the package, so its own share is unknown.

In 2014 two open letters showed how divided scientists were. One, organised by the Stanford Center on Longevity and the Max Planck Institute for Human Development and signed by about 70 researchers, said that the claims made for brain games were not supported by the evidence. A reply signed by over a hundred others argued that some forms of training do have measurable benefits. Both sides agreed on one thing: the marketing was ahead of the science.

The dementia question

The most hopeful single result comes from a 10-year follow-up of the ACTIVE trial, which reported that people who did the speed-of-processing training were diagnosed with dementia less often than the control group (a reduction of about 29 percent). It was a secondary analysis, not the question the trial was designed to answer, and it has not been confirmed by another trial. So far, no brain game has been shown to prevent dementia.

The Lumosity case

In January 2016 the US Federal Trade Commission charged Lumos Labs, the company behind Lumosity, with deceptive advertising. Its ads had said the games could improve performance at work and school and delay age-related cognitive decline, including in conditions such as dementia. The FTC said these claims were not backed by science. The company settled for 2 million dollars (from a 50 million dollar judgment that was suspended because of its finances) and agreed to stop making such claims without reliable evidence.

Why skills do not transfer easily

The brain learns specifically. When you practise a Schulte table, you get faster at moving your eyes across a grid, ignoring the numbers you have already found and keeping your place. Those are real skills, but they are narrow. They do not automatically make you better at reading a contract or remembering a phone number, because those tasks use different knowledge and strategies.

Studies of experts show the same pattern. Chess masters can recall a real game position after a glance, but when pieces are placed at random they remember them little better than beginners: their memory advantage comes from knowledge of chess, not from a stronger memory in general. Meta-analyses by Giovanni Sala and Fernand Gobet found that chess and music training also show little far transfer to school results or intelligence. Working memory itself is small: most adults can hold about four meaningful items in mind at once (George Miller's famous "seven, plus or minus two" from 1956 was later revised downwards by Nelson Cowan and others), and practice changes how cleverly you use that space more than its size.

What does help the brain

If the goal is a healthier brain over the years, the evidence points to habits outside the screen. The Lancet Commission on dementia (2024 update) lists 14 modifiable risk factors and estimates that nearly half of dementia cases are linked to them. Among them:

  • Physical activity. In a 2011 study by Kirk Erickson and colleagues, older adults who walked briskly for a year saw the hippocampus, a brain region central to memory, grow by about 2 percent, while it shrank in the stretching group.
  • Hearing and vision. Untreated hearing loss in midlife is one of the largest single risk factors in the Lancet report; hearing aids are a brain intervention as much as an ear one.
  • Blood pressure, diabetes, cholesterol and smoking. What is good for the heart's blood vessels is good for the brain's.
  • Education and staying mentally and socially active. Learning something demanding and new, such as a language, an instrument or a craft, combines memory, attention, motor skill and social contact in a way short games cannot.
  • Sleep. Memories are consolidated during sleep; a short night costs more attention than any game can win back.

So are brain games worth playing?

Yes, as long as you expect the right things. A good puzzle or reaction game gives you a few minutes of full concentration, a clear goal and visible progress, and that is pleasant and useful in itself. For older players, a daily game can be a gentle way to keep a routine and notice changes. For anyone, it is a better break than scrolling. Here is how to get the most out of them:

  1. Vary the games. Since gains stay close to what you train, several different tasks cover more ground than one.
  2. Keep it hard. Once a level feels automatic, raise the difficulty. Learning happens at the edge of what you can do.
  3. Keep sessions short. 10 to 20 minutes is plenty. Longer sessions mostly train patience.
  4. Track your own times, not a "brain age". Scores within one game are meaningful; a single number for your whole brain is not.
  5. Combine it with real life. A walk, a conversation or a new hobby does what a game cannot.

What our games train (and what they do not)

All the games on Extensive Harmony are free, run in the browser and need no account. None of them claims to raise your IQ. Here is what each one actually exercises:

  • Schulte Table: visual search and sustained attention. You find the numbers in order on a 4×4, 5×5 or 6×6 grid; there is no time limit, only your own time to beat.
  • Color Echo: short-term memory for sequences. Each round adds one step; on Easy every colour stays lit for 0.9 seconds, on Medium for 0.55 seconds. Our Color Echo memory tips show how chunking extends your record.
  • Pairs: visuospatial memory, with 8, 12 or 18 pairs. Our simulation of 1.5 million games found that a perfect memory needs about 1.6 tries per pair, a useful target (see the maths of memory games).
  • Quick Math: arithmetic fluency under time pressure, with 15, 10 or 7 seconds per question. Unlike most brain games, this one trains a skill you use every day.

For calmer games without a clock, see free brain games for seniors.

Red flags in brain-training ads

  • Promises to raise IQ, prevent Alzheimer's or improve school or work results.
  • A single "brain age" or "brain score" presented as a medical measure.
  • "Clinically proven" with no link to a published, controlled study.
  • Studies where the control group did nothing, instead of doing another activity: then any gain could come from being busy and motivated.
  • Subscriptions sold with urgency, as if every missed day were damage.

Frequently asked questions

Does brain training actually work?

Partly. Brain training reliably makes you better at the trained task and at very similar tasks. Large reviews have found little evidence that it raises general intelligence, improves everyday memory or prevents dementia. Treat brain games as an enjoyable way to practise focus, not as a medical treatment.

Can brain games prevent dementia?

There is no proof that they do. One long-term trial, ACTIVE, reported a lower dementia rate after speed-of-processing training, but it was a secondary analysis and has not been confirmed. Physical activity, treating hearing loss, high blood pressure and diabetes, not smoking and staying socially and mentally active have stronger evidence.

What is near transfer and far transfer?

Near transfer means getting better at tasks very similar to the one you trained, for example a different grid of numbers after practising Schulte tables. Far transfer means improving in something quite different, such as general intelligence or remembering names. Brain training shows near transfer reliably and far transfer rarely.

Why was Lumosity fined?

In January 2016 the US Federal Trade Commission charged Lumos Labs, the maker of Lumosity, with deceptive advertising for claiming its games could improve performance at work and school and delay age-related decline without scientific support. The company agreed to pay 2 million dollars to settle.

How long should I play brain games each day?

There is no proven ideal dose. Short sessions of 10 to 20 minutes on several different games, at a level that feels challenging but not frustrating, are a sensible habit. If a game stops being hard, raise the difficulty or switch to something new.