Guide11 min read
Does brain training work?
What the research shows, where it works best, and how to train so it counts.
The short answer
Yes. Brain training reliably makes you better at the skills you practise, and those gains often spread to similar tasks.
How far they spread depends on what you train. Spatial training has some of the strongest evidence: the gains last, and they carry over to spatial tasks people never practised.
Bigger promises, such as a large jump in IQ, have mixed evidence. But the research is clear about what gives you the best chance: training that keeps getting harder, covers more than one skill, and fits into a regular routine.
What brain training is
Brain training means practising mental tasks on purpose to get better at the thinking skills behind them. Today most of it happens on a screen, as short games that get harder as you improve.
Most brain training works on one of five skills:
- Working memory: holding and updating information from moment to moment, as in dual n-back.
- Reasoning: working out how things relate and what follows, as in relational frame training.
- Spatial skill: picturing shapes and turning them in your head, as in mental rotation training.
- Divided attention: following several moving things at once, as in multiple object tracking.
- Processing speed: scanning a busy scene and picking out what matters quickly.
Your brain changes with practice
Brain training rests on a simple idea: the adult brain keeps changing when you ask it to learn. Two well-known studies show this clearly.
Eleanor Maguire and colleagues (2000) scanned the brains of London taxi drivers, who spend years learning the city's streets. The back part of the hippocampus, an area used for spatial memory, was larger in the drivers than in people who didn't drive taxis. The longer someone had driven a taxi, the larger it was.
Bogdan Draganski and colleagues (2004) taught people to juggle for three months. Brain scans showed more grey matter in areas that handle seeing movement. When people stopped practising, some of that gain faded.
Neither study used brain games. But both show that the brain responds to practice, and that keeping it up matters.
Near and far transfer
Transfer is the most important word in brain training research. It means a skill you practised helping with something you didn't practise.
Near transfer is a gain on tasks close to the one you trained, such as a different memory game after training on a memory game. Far transfer is a gain on something quite different, such as reasoning tests, school results or everyday life.
- Daily lifeSchool, work and everyday tasksEvidence: Still being studied
- Broad abilitiesReasoning and IQ-style testsEvidence: Mixed
- Similar tasksOther tasks that use the same skillEvidence: Strong
- The task you trainGetting better at the game itselfEvidence: Very strong
This is the question to ask of any brain training claim: which rung is it talking about?
The lower rungs matter more than they might sound. Near transfer means the skill itself has improved, not just your score on one game. And as the rest of this guide shows, some kinds of training reach further up than others.
What the research says
Brain training research has swung between excitement and doubt. The debate has made the science sharper, and it's worth knowing how it got here.
The early excitement
In 2008, Susanne Jaeggi and colleagues reported that training on dual n-back raised scores on tests of fluid intelligence, the kind of reasoning IQ tests measure. It made headlines and helped launch a wave of brain training apps.
Our dual n-back guide covers that study and the replications that followed.
The pushback
In 2010, Adrian Owen and colleagues ran a six-week online study with 11,430 people. They got better at every task they trained, but not at untrained tests, even ones close to the training.
People were asked to train for at least 10 minutes a day, three days a week. That is a light dose next to most training studies, which is one reason researchers still debate what the study shows.
In 2014, two groups of scientists published open letters that disagreed. More than 70 said brain games were not a proven way to improve thinking. Months later, 133 others said there was plenty of evidence that they helped.
In 2016, the US Federal Trade Commission fined one well-known brain training company $2 million over ad claims it could not back up. Brain training itself was not the problem. Overselling it was.
The same year, Daniel Simons and colleagues published the largest review of the field so far. They found extensive evidence that training improves the trained tasks, less evidence for closely related tasks, and little for distant tasks or everyday life. They also found that many studies had design weaknesses.
Where things stand now
Lampit, Hallock and Valenzuela (2014) pooled randomised trials of computer brain training in healthy older adults.
52 trialsPooled by Lampit et al. (2014), covering 4,885 healthy older adults. Training gave a small but reliable boost to thinking skills.
The clearest gains were in processing speed, spatial skills and memory. How the training was set up made a big difference, which is covered below.
Not everyone reads the evidence so positively. Sala and colleagues (2019) combined earlier reviews. They agreed that near transfer is real, but concluded that far transfer disappears once placebo effects and unpublished null results are taken into account.
Katz, Shah and Meyer (2018) argue that "does brain training work?" is too blunt a question. Studies differ so much in how they are designed that pooling them can hide what's going on.
The more useful question is which training works, for whom, and under what conditions. The next section looks at where the answers are most encouraging.
Where the evidence is strongest
Brain training isn't one thing, and the evidence differs from one type to the next. Here is where it looks best.
Spatial skill
Uttal and colleagues (2013) combined 217 spatial training studies. Training improved spatial skill by about half a standard deviation on average, a medium-sized effect.
217 studiesPooled by Uttal et al. (2013). Spatial training gains lasted, and spread to spatial tasks people never practised.
That is near transfer at its best: gains that stick and carry over. Spatial skill also has a long record of predicting success in science and engineering. Our mental rotation training guide covers both, with a free version of the classic test to try.
Working memory
Dual n-back is the most studied working memory task. Soveri and colleagues (2017) pooled 33 randomised trials. People made medium-sized gains on n-back tasks they hadn't trained, and smaller gains on other memory, attention and reasoning tests.
Au and colleagues (2015) estimated an average gain of 3 to 4 IQ points across 20 n-back studies. Other researchers found smaller effects in the strictest studies.
Your level matters. Pahor, Seitz and Jaeggi (2022) found that the people who improved most on n-back were the ones who improved most on reasoning. Our dual n-back guide explains why pushing to higher levels counts.
Relational reasoning
Relational frame training teaches you to work out relations nobody told you. If A is bigger than B, and B is bigger than C, is A bigger than C? Early studies reported large IQ gains in children and teenagers.
A 2022 review found a moderate effect on nonverbal IQ across five trials, but rated all five at high risk of bias. In a 2026 trial with older adults, relational training and dual n-back improved executive function by about the same amount.
It's a young field with promising early results. Our relational frame training guide goes through every study.
Divided attention
Multiple object tracking asks you to follow several moving targets at once. In a 2024 study, 58 college basketball players were split into three groups. After eight sessions of tracking training, both trained groups improved on tests of executive function, and the group that got instant feedback improved most.
The studies so far are small, but the early signs are good.
At a glance
| Training type | What the studies show | Evidence | Read more |
|---|---|---|---|
| Spatial skill | Medium gains that last and spread to spatial tasks people never practised | Strong | Mental rotation guide |
| Working memory | Big gains on the task, smaller ones beyond it, and more for people who keep progressing | Growing | Dual n-back guide |
| Relational reasoning | Striking early results, with more mixed findings in stricter studies | Early but promising | Relational frame training guide |
| Divided attention | Gains on attention and memory tasks in small studies of athletes | Early but promising | Orbit |
What makes training work
The studies that found the biggest gains have a few things in common. These are the ingredients worth looking for.
It keeps getting harder
A level you have mastered stops being a workout, and a level far above you turns into guessing. The useful zone is just past what feels easy.
That matters for more than the task itself. In the n-back studies, the people who kept improving were the ones who showed wider gains. Training that adapts to you keeps you in that zone without you having to think about it.
It covers more than one skill
Gains stay closest to what you train. So the surest way to broad benefit is to train a range of skills, rather than one game over and over.
It fits into a routine
In Lampit's review of older adults, training three times a week or less did better than training more often. Training in a supervised group also did better than training alone at home.
Other studies trained people more often, so treat the first finding as a sign that rest matters rather than a strict rule. Our suggestion is short sessions a few times a week, kept up for weeks.
Training at home can't copy a supervised group, but a fixed routine and a clear record of your progress give you some of the same structure.
It sits alongside healthy habits
The FINGER trial gave 1,260 people aged 60 to 77, all at higher risk of memory decline, two years of exercise, diet advice, brain training and health checks. They did better on thinking tests than a group given general health advice.
Brain training does most for you as one part of a healthy routine, alongside moving more, eating well and sleeping well.
Myths and facts
| Myth | What the evidence shows |
|---|---|
| Brain training will make you a genius. | No study backs a huge IQ jump. What you can expect is to get sharper at the skills you train, and sometimes a little beyond them. |
| Brain training is a scam. | Training reliably improves the skills you practise, and some kinds show lasting, wider gains. Some ads have overclaimed, but the training itself works. |
| Any puzzle does the same job. | Gains stay closest to what you practise, so a daily sudoku mostly makes you better at sudoku. Training a range of skills covers more ground. |
| You're too old to benefit. | The adult brain keeps changing with practice, and across 52 trials training gave healthy older adults a small but reliable boost. |
| Once you improve, it lasts forever. | Spatial training gains held up well over time, but skills you stop using can fade. Keep practising to keep your gains. |
Honest brain training promises what the research supports: sharper skills, built through steady practice. That is worth having, and it's what we built Neurobuff around.
How to spot good brain training
Not every brain training app is built the same way. Look for these five things:
- It gets harder as you improve.Training that adapts keeps you just past what feels easy. In the studies, the people who kept improving were the ones who showed wider gains.
- It trains more than one skill.Gains stay closest to what you practise, so a range of skills covers more ground than one game.
- It is based on published research.The tasks should come from studies you can look up. Good programs tell you which ones.
- It doesn't promise IQ points or a cure.No study backs a big IQ jump, and no brain game is proven to prevent dementia. Promises like these are a warning sign.
- It shows your progress.A clear record of how you are doing helps you build a routine and see that the practice is paying off.
A program that ticks all five follows what the research says works best. We built Neurobuff around them.
How Neurobuff does it
Neurobuff has five training games, each built on a different line of research, so you train a range of skills rather than one.
Shapes come out of nine portals while a voice reads letters. Portals
Read the dual n-back guide
Short premises in made-up words, then a Yes or No question. Terminal
Read the relational frame training guide
A cube structure to rebuild from a new angle. Blocks
Read the mental rotation guide
Several moving targets to follow as they cross paths. Orbit
See Orbit on the homepage
A busy scene to scan for what matters, fast. Observatory
See Observatory on the homepage
Every game sets its difficulty from how you are doing, so each session stays just past what feels easy. As you improve, the games move you on to harder levels.
Short daily challenges and a record of your progress help you build a routine. The Academy covers the habits that support your brain outside the games, such as exercise, food and sleep.
Ready to try it? Start training and play all five games.
Frequently asked questions
Does brain training actually work?
Yes, for the skills you practise. Studies consistently find that people get better at the tasks they train, and those gains often carry over to similar tasks.
How far the gains spread depends on the type of training. Spatial training has some of the strongest evidence. Claims of big jumps in IQ go further than the research does.
Is brain training worth it?
For most people, yes, as long as you know what to expect. You can expect to get sharper at the skills you train, and some kinds of training, such as spatial training, give gains that last and spread to similar tasks.
Choose training that adapts to you and covers several skills, and treat it as one part of a healthy routine.
Is brain training a scam?
No. The training itself does something real: it improves the skills you practise, and some kinds show lasting, wider gains.
Some companies have overstated what it can do. In 2016, the US Federal Trade Commission fined one well-known brain training company $2 million over claims its ads could not back up.
Can brain training increase your IQ?
The evidence is mixed. A 2015 review of 20 n-back training studies estimated an average gain of 3 to 4 IQ points. Other researchers found smaller effects in the best controlled studies.
Several studies found that the people who improve most on the training task are the ones who gain most on reasoning tests. That makes it worth pushing to harder levels.
Can brain training improve memory?
Yes, for the kinds of memory you practise. A review of 52 trials with healthy older adults found small gains in several kinds of memory, including working memory.
N-back training gives clear gains on memory tasks like the one you train, and smaller gains on other memory tasks.
Does brain training help with focus?
It can help with the attention skills you practise. In a 2024 study, basketball players who trained on multiple object tracking got better at a test of holding back quick responses. A review of n-back studies found smaller but real gains in attention control.
Evidence that this carries over to focus in everyday life is still limited.
Can brain training prevent dementia?
No brain game has been proven to prevent dementia, and Neurobuff does not claim to.
The most encouraging evidence is for healthy habits together. In a two-year trial, older adults at higher risk who combined exercise, a healthy diet, brain training and health checks did better on thinking tests than a group given general health advice.
If you are worried about your memory, speak to a doctor.
How long does brain training take to work?
You will usually get better at a game within your first few sessions. Most studies of wider gains ran for several weeks.
Spatial training gains held up when people were tested again later, so the effort tends to last.
What is the best type of brain training?
It depends on what you want to improve, because gains stay closest to the skill you train. Spatial training has some of the strongest evidence so far.
Training several different skills gives you the widest coverage, which is why Neurobuff has five games, each built on a different line of research.
Are brain games better than crosswords and sudoku?
They do different jobs. Crosswords and sudoku are good practice for word and number puzzles, and because gains stay closest to what you practise, they mostly make you better at those puzzles.
Adaptive brain games are built on tasks from training studies. They target skills such as working memory and spatial skill, and get harder as you improve. There is no harm in doing both.
How often should you do brain training?
Short sessions a few times a week are a good start. In a review of studies with healthy older adults, training three times a week or less worked better than training more often.
Keeping it up for weeks matters more than doing long sessions.