Crossword puzzles help some things and not others, and the difference matters. They are excellent at building vocabulary and word retrieval. They are linked, in observational studies, to slower memory decline. But the headline "crosswords prevent memory loss" oversimplifies a more interesting and more honest picture.
This post separates what the research actually supports from what gets repeated without evidence.
The short answer. Crosswords reliably train word knowledge and retrieval. The best controlled trial to date found crosswords outperformed computer brain games in people with mild cognitive impairment. But observational links between puzzles and slower decline are partly explained by reverse causation — sharper people do more puzzles. Crosswords are a genuinely good habit; they are not a cure.
What crosswords actually train
A crossword is a word-retrieval task with a logic scaffold. You hold a clue in working memory, search long-term memory for candidates, and check each against the letters already on the grid. Done regularly, this builds two things well: a larger accessible vocabulary, and faster, more reliable retrieval of words you already know.
That is a real cognitive benefit. The "tip of the tongue" experience — knowing a word is there but not being able to reach it — is a retrieval problem, and retrieval practice is exactly what crosswords drill. If word-finding is your specific complaint, crosswords are well matched to it.
What crosswords do not heavily train is working-memory capacity, processing speed, or sustained attention under interference. Those are different abilities, exercised by different tasks.
What the observational studies show
Several large studies have linked mentally engaging leisure activities to better cognitive outcomes:
- Verghese et al., 2003 (NEJM). Following older adults in the Bronx Aging Study, researchers found that frequent participation in cognitively engaging leisure activities — including crossword puzzles, reading, and playing board games — was associated with a lower risk of dementia.
- Pillai et al., 2011. Among people who eventually developed dementia, those who had done crossword puzzles regularly showed a delay in the onset of accelerated memory decline.
These findings are consistent with the cognitive-reserve hypothesis: a lifetime of mental engagement appears to help the brain tolerate age-related changes for longer. But observational studies share one stubborn limitation — reverse causation. People with better memory and verbal ability are more likely to enjoy and keep doing crosswords. Some of the apparent benefit is the puzzle reflecting an already-healthy brain, not building one.
The strongest controlled evidence
The most rigorous test came in 2022. Devanand and colleagues, publishing in NEJM Evidence, randomized 107 adults with mild cognitive impairment to either web-based crossword puzzles or computerized cognitive games, over 78 weeks.
The result surprised many people: the crossword group did better. They showed greater improvement on a standard cognitive scale, less brain shrinkage on MRI, and better day-to-day function than the computer-games group. This is the cleanest evidence available that crosswords are not merely a marker of a healthy brain — for this population, they appeared to do something.
Two cautions keep this honest. The study was relatively small, and it was specific to people who already had mild cognitive impairment. It does not prove crosswords prevent decline in healthy adults, and it does not settle the broader debate.
Crosswords versus targeted brain training
It is tempting to read the 2022 trial as "crosswords beat brain-training apps, case closed." It is not that simple. That trial tested one set of computer games, not the targeted training the wider literature supports.
The 2024 Sala and Gobet meta-analysis found that targeted training — working memory, processing speed, attention — produced modest but real benefits, while generic puzzle games did not. Crosswords and targeted training are not interchangeable. They exercise different abilities:
| Activity | Trains well | Trains weakly |
|---|---|---|
| Crossword puzzles | Vocabulary, word retrieval | Working memory, processing speed |
| Targeted brain training | Working memory, attention, speed | Vocabulary, general knowledge |
The useful framing is not "which is better" but "which ability do you want to work on." If word-finding is the problem, crosswords. If holding information in mind and resisting distraction is the problem, targeted training is the better-matched tool.
Where crosswords fit in brain health
The 2024 Lancet Commission on dementia identified 14 modifiable risk factors across the lifespan — including hearing loss, hypertension, physical inactivity, social isolation, and depression. Mentally and socially engaging activity fits the broader pattern the Commission supports, but no single puzzle appears on that list as a standalone intervention.
That is the right scale to think at. Crosswords are one good habit. They sit alongside physical exercise, good sleep, social connection, and managing blood pressure — and none of those, alone, is a guarantee.
The honest takeaway
Do crossword puzzles help memory? Yes, in a specific and worthwhile way: they build vocabulary and word retrieval, the best controlled trial found a real benefit in people with mild cognitive impairment, and regular mental engagement is consistently linked to slower decline.
But they are not a memory cure, and they do not train every ability that matters. If you enjoy crosswords, keep doing them — enjoyment is what makes a habit last. If your concern is the broader set of everyday memory skills — remembering names, holding a list, staying focused — pair them with practice built around exactly those tasks. That is what we designed BrightYears to do: a short daily memory practice for adults 45+, in five focused minutes a day.