The cognitive benefits of puzzles: what the research really says

The short answer

Doing puzzles is a genuinely good use of your time. They are absorbing, they make you think, and people who do them regularly tend to score better on tests of memory and reasoning. What the research can’t yet tell us is how much of that difference puzzles actually cause, and whether getting better at Sudoku makes you better at remembering names or managing your finances.

If you only remember four things from this guide, make them these:

• You improve at what you practice. This is the most consistent finding in the field.
• Transfer to other skills is limited. Gains rarely spread far beyond the trained task.
• Associations are not proof. Sharper people may simply enjoy puzzles more.
• Puzzles are one part of a bigger picture. Physical activity, social contact, sleep and heart health all matter for the aging brain.

What happens in your brain when you solve a puzzle

A puzzle is a small, well-defined problem with a clear goal. Solving one draws on a group of mental abilities that psychologists often call executive functions and related skills:

• Working memory: holding a few pieces of information in mind while you use them, such as the candidates for a Sudoku cell or the letters you have already tried in a word game.
• Attention: staying focused on the task and filtering out distractions.
• Reasoning: working out what must be true from the rules and the clues.
• Processing speed: how quickly you take in information and respond, which matters most in timed and arcade games.
• Retrieval: pulling words, facts and patterns out of long-term memory, the core of crosswords and trivia.
• Visuospatial skills: mentally rotating, fitting and comparing shapes, as in jigsaws and block puzzles.

These abilities tend to change with age at different rates. Speed and working memory generally begin to slow gradually in adulthood, while vocabulary and general knowledge often hold steady or keep growing into later life. That is one reason word puzzles can feel easier with age while fast-paced arcade games feel harder.

“Use it or lose it” and cognitive reserve

The popular idea behind puzzles for brain health is “use it or lose it”: keep the mind busy and it will stay in better shape. It is an intuitive idea, and the brain does change with experience. But psychologists who have tested the hypothesis carefully, including the cognitive aging researcher Timothy Salthouse in an influential 2006 review, have pointed out that the evidence that mental activity slows the rate of age-related decline is weaker than the slogan suggests.

A more useful concept is cognitive reserve, developed by the neuroscientist Yaakov Stern and others. It describes the observation that people with similar amounts of age-related change or disease in the brain can show very different levels of everyday function. Factors such as education, a mentally demanding job, and a lifetime of engaging leisure activities are associated with more reserve. The idea is not that these activities stop brain changes, but that they may help the brain cope with them for longer.

Puzzles fit naturally into the “engaging leisure activities” part of that story. What remains uncertain is whether taking up puzzles later in life builds reserve in a meaningful way, or whether the benefit comes from a lifetime of curiosity and learning of which puzzling is just one sign.

What the research actually shows

Large observational studies: the PROTECT puzzle findings

In 2019, researchers from the University of Exeter and King’s College London published two papers in the International Journal of Geriatric Psychiatry using data from the PROTECT online study, which included more than 19,000 adults aged 50 and over in the UK. Participants reported how often they did word puzzles, such as crosswords, and number puzzles, such as Sudoku, and completed a battery of online cognitive tests.

People who reported doing these puzzles more often tended to perform better on tests of attention, reasoning and memory, and the relationship was graded: more frequent puzzlers tended to score higher. That is an encouraging pattern. It is also exactly the kind of result that can’t tell us which way the arrow points. People whose thinking is already sharp may be more likely to enjoy and stick with puzzles, and the researchers themselves described the findings as associations that needed testing in trials.

Crossword puzzles and later decline

Other long-running studies of older adults, such as the Bronx Aging Study in New York, have reported that people who did crosswords regularly showed a later onset of accelerated memory decline among those who went on to develop dementia. Again, the design is observational, so it fits several explanations, including the possibility that reserve delays the point at which symptoms show.

Randomized trials: the ACTIVE study

The strongest evidence on cognitive training comes from randomized controlled trials, where participants are assigned to training or a comparison group by chance. The best known is ACTIVE (Advanced Cognitive Training for Independent and Vital Elderly), a US trial that began in the late 1990s with nearly 3,000 adults aged 65 and older.

Participants received about ten sessions of training in one of three areas: memory strategies, reasoning (spotting patterns in sequences) or speed of processing (a computerized task requiring quick visual attention), with some receiving later booster sessions. Each type of training improved the specific ability it targeted, and some of those gains were still measurable years later. Follow-ups over a decade also reported that trained groups described somewhat less difficulty with some everyday activities than the control group, and a later analysis suggested speed-of-processing training was associated with a lower risk of dementia diagnosis. Those longer-term findings drew wide attention but were secondary analyses that need independent confirmation.

Two cautions apply. First, ACTIVE tested structured training programs, not crosswords or Sudoku, so its results don’t translate directly to casual puzzling. Second, training mostly improved the trained skill rather than cognition across the board.

Crosswords versus brain-training games

A small randomized trial published in NEJM Evidence in 2022 compared web-based crossword puzzles with web-based computerized cognitive games in older adults with mild cognitive impairment over about a year and a half. The crossword group did modestly better on the main cognitive measure. It is an interesting result that suggests familiar puzzles hold their own against purpose-built software, but the study was small, had no no-activity comparison group and needs to be replicated before firm conclusions are drawn.

Jigsaw puzzles

A 2018 study from Germany found that jigsaw puzzling draws on many visuospatial abilities at once, and that lifetime jigsaw experience was associated with better visuospatial cognition. In the same project, however, a month of regular puzzling did not produce measurable cognitive gains compared with a control group. Short bursts of any single activity seem unlikely to move broad measures of cognition.

Brain-training claims, and the Lumosity case

The commercial brain-training boom of the 2000s and 2010s promised a lot. In 2014, a group of around 70 scientists coordinated by the Stanford Center on Longevity and the Max Planck Institute for Human Development issued a statement warning that claims about brain games preventing or reversing cognitive decline were not supported by the evidence. Another group of researchers responded with a letter arguing that some forms of training showed real promise. Both sides agreed that more rigorous research was needed.

A major 2016 review in Psychological Science in the Public Interest, led by the psychologist Daniel Simons, examined the studies cited by brain-training companies. It found extensive evidence that training improves performance on the trained tasks, less evidence for improvement on closely related tasks, and little evidence that training improves everyday cognitive performance.

The same year, the US Federal Trade Commission settled charges against Lumos Labs, the company behind Lumosity, over advertising that claimed its games could improve performance at work and school and delay age-related decline. The FTC said those claims were not backed by science. The company agreed to pay $2 million and to stop making such claims without adequate evidence.

That is why we don’t promise that our brain training games will raise your IQ or protect against dementia. They are enjoyable challenges that exercise specific skills, and that is a perfectly good reason to play them.

Limits of the evidence

• Reverse causation. Good cognition may lead to puzzling rather than the other way round.
• Confounding. Puzzlers may differ in education, income, health and social life, all of which affect cognition.
• Self-report. Many studies rely on people estimating how often they do puzzles.
• Narrow transfer. Trials usually show gains on the practiced task, with less spread to other abilities.
• Small or short trials. Many studies are too small or brief to detect effects on real-world outcomes like dementia.

Health organizations reflect this caution. The World Health Organization’s 2019 guidelines on reducing the risk of cognitive decline and dementia concluded that the evidence for cognitive training was of low certainty, while giving stronger weight to factors like physical activity, not smoking and managing conditions such as high blood pressure and diabetes. The Lancet Commission on dementia prevention has likewise highlighted modifiable risk factors including physical inactivity, hearing loss, social isolation and less education early in life.

Which skills different puzzles exercise

Because gains are largely specific, variety is your friend. Here is a practical map of what our games ask of you:

A sensible puzzle routine for a sharper mind

Nobody can give you a proven prescription, but the research does point to some sensible principles: keep it challenging, keep it varied, keep it enjoyable enough that you continue, and keep it in proportion with the rest of a healthy life.

1. Start with one daily puzzle. Our daily challenge gives you a fresh puzzle each day, the same one everyone else is playing. A fixed daily slot, such as with your morning coffee, is easier to keep than good intentions.
2. Add a second skill. If your daily puzzle is a word game, follow it with a logic or spatial puzzle, or the other way round. Aim for 10 to 20 focused minutes in total.
3. Stay at the edge of your ability. Once a difficulty level feels automatic, step up. Learning happens when you have to work for the answer, not when you’re on autopilot.
4. Try something unfamiliar each week. Learning a new game’s rules and strategies is itself mentally demanding. Browse the new games page or pick a category you usually skip.
5. Reflect, don’t just repeat. After a hard puzzle, notice the technique that cracked it. Our guides explain strategies for most games so you can learn why a method works.
6. Make it social. Compare daily results with friends or climb the leaderboards. Social connection is valuable for brain health in its own right.
7. Balance it. A walk, a conversation and a good night’s sleep are at least as important as another grid.

If you are noticing changes in your memory or thinking that worry you, please talk to a doctor. Puzzles are a pleasure, not a diagnosis or a treatment. For the broader, everyday reasons people love puzzles, from stress relief to family game time, see our companion guide to the benefits of puzzles.

Frequently asked questions

This guide is general information, not medical advice. We review it periodically as new research is published.