Neither is a proven way to protect your memory, so the better choice is the one you will enjoy and keep doing. Crosswords mostly draw on vocabulary and general knowledge, cost little or nothing, and in one small 2022 trial of people with mild cognitive impairment they did slightly better than computerized brain games. Brain training apps exercise a wider range of skills, such as speed, attention, and working memory, adjust their difficulty as you improve, and cost $39.99 to $96 a year for the two apps we recommend below, though free tiers exist. Many people do best with some of both.
The daily crossword has been recommended for brain health for generations, and brain training apps promise something more targeted. This guide compares the two directly. For the broader question of whether brain training works at all, see our guide on whether brain training apps actually work.
Crosswords vs brain training apps at a glance
| Crosswords | Brain training apps | |
|---|---|---|
| Main skills used | Vocabulary, general knowledge, word retrieval | Processing speed, attention, working memory, reasoning; some practice reading and math |
| Difficulty | You pick the puzzle | Adjusts automatically as you improve |
| Cost | Free to low cost | Free tiers; $39.99 to $96 a year for our two picks |
| Format | Paper or screen | Phone, tablet, or computer |
| Social | Easy to share or solve together | Mostly solo |
| Head-to-head evidence | Beat computerized games in one small trial of people with MCI | Beat crosswords in one large online trial run by Lumosity’s maker |
What crosswords actually exercise
Crosswords lean heavily on what psychologists call crystallized knowledge: the vocabulary and facts you have built up over a lifetime. In four studies of about 200 adults each, aged 18 to 80, general knowledge was the strongest predictor of how good people were at crosswords, and abstract reasoning ability had no direct link to puzzle skill (Hambrick, Salthouse and Meinz, 1999).
The same research found no evidence that years of crossword experience slowed the age-related decline in fluid abilities such as reasoning. Crystallized knowledge tends to grow with age, which helps explain why many older solvers are so good at crosswords, but it also means a regular solver is mostly practicing a strength. To make crosswords stretch you, move up a difficulty level, try a new style such as cryptic crosswords, or add a number puzzle like sudoku.
What the long-term studies show
The best-known evidence for crosswords is observational. The Bronx Aging Study followed 488 older adults who started out without cognitive problems. Among the 101 who later developed dementia, those who did crossword puzzles at the start of the study began their period of rapid memory decline about 2.5 years later than those who did not (Pillai et al., 2011).
That is an association, not proof. People who keep doing puzzles may differ in many other ways, and the finding concerns the timing of decline, not whether crosswords prevent dementia. The authors called for clinical trials.
What brain training apps exercise
Apps such as BrainHQ, Elevate, and Lumosity use short games aimed at specific abilities: processing speed, attention, working memory, reasoning, and, in Elevate’s case, reading, writing, and math. The difficulty adapts as you improve, and most apps chart your progress.
In brief, you reliably get better at the games you practice, but evidence that those gains carry over into everyday life is limited (Simons et al., 2016). The strongest results are for speed-of-processing training in older adults from the ACTIVE trial, and the National Institute on Aging says there is not enough evidence that commercial apps have the same impact. Our evidence guide covers the caveats in full.
The head-to-head research
Two randomized trials have compared brain games directly with crosswords, and they point in opposite directions.
The 2022 trial: crosswords beat computerized games
The trial behind the headlines, published in NEJM Evidence (Devanand et al., 2022) and funded by the National Institute on Aging, enrolled 107 adults aged 55 to 95 (average age 71) with mild cognitive impairment (MCI) at Columbia and Duke. They were randomly assigned to web-based crosswords or web-based cognitive games, both provided free by Lumos Labs, the maker of Lumosity, which had no role in the study design or interpretation. Both groups trained at home for 30 minutes, four times a week, for 12 weeks, then did six week-long booster periods across 18 months.
What it found:
- On the main measure, a cognitive test used in Alzheimer’s research that is scored from 0 to 70 (higher is worse), the crossword group’s average score improved slightly, from 9.59 to 8.61, while the games group’s average worsened slightly, from 9.53 to 9.93.
- The crossword group also did better on a questionnaire about everyday functioning.
- Brain scans showed less shrinkage of the hippocampus and less thinning of the cortex in the crossword group.
- Other secondary measures, including a broader battery of thinking tests, showed no difference.
- Slightly fewer people dropped out of the crossword group (12.5%) than the games group (17.6%).
The caveats, many from the authors themselves:
- There was no untrained group, so the study cannot show whether either activity helped compared with doing neither. The authors note both might have looked beneficial against a control group.
- It was small and specific: 107 people with MCI, most of them highly educated, so it may not apply to people without memory problems.
- The games were general cognitive games aimed at memory and executive skills, not the speed-of-processing training with the strongest ACTIVE results.
- It needs repeating. The authors write that no single trial is definitive.
The authors offer one explanation: people with more advanced MCI may have found the games too hard to follow, while most older adults already know crosswords, which in the trial were of medium difficulty and self-paced.
The 2015 trial: Lumosity beats crosswords
A much larger online trial went the other way. It randomly assigned 4,715 adults to either Lumosity’s training program of 49 exercises or crossword puzzles, for about 15 minutes at least five days a week over 10 weeks (Hardy et al., 2015). The Lumosity group improved more on a combined set of cognitive tests and on self-rated concentration, with small effects.
The major caveat is who ran it: most of the authors worked for Lumos Labs and held stock options in the company. Testing was online, follow-up was short, and the outcomes were cognitive tests rather than everyday life.
What the two trials suggest together
Neither trial shows that one activity protects your brain. The 2022 authors suggest the answer may depend on who is doing it: games may work better for healthy adults, especially younger people used to games, while familiar, less technical crosswords may suit older people with cognitive impairment. The practical lesson is that an engaging activity you do consistently matters more than the perfect format.
Cost
Crosswords can cost almost nothing: the newspaper you already read, a puzzle book, or a free daily puzzle online. Our guide to daily puzzle games like Wordle covers the main puzzle apps, and our list of games like Connections adds more word-grouping puzzles.
Brain training apps have free tiers, but the full versions cost more. Elevate is $39.99 a year on the App Store. BrainHQ is $96 a year or $14 a month on its website, though BrainHQ says some Medicare Advantage plans and many US public libraries offer it at no cost. Our best free brain training apps guide covers the free tiers.
Enjoyment and sticking with it
Any benefit depends on doing it regularly, so enjoyment is not a side issue. Crosswords have an easy social side: you can solve with a partner or talk through a tricky clue. Apps offer streaks, scores, and short sessions that fit into spare minutes.
Challenge matters too. In the Synapse Project, older adults who spent three months learning a demanding new skill, quilting or digital photography, improved their memory more than people who did easy mental tasks or simply socialized (Park et al., 2014). A puzzle you finish on autopilot asks less of you, so whichever you choose, keep raising the difficulty.
Who should do which
Choose crosswords if you enjoy words and trivia, you want something low-cost or on paper, you like solving with other people, or you would rather spend less time on screens.
Choose a brain training app if you want variety beyond words, such as speed, attention, and numbers, you like difficulty that adjusts itself and progress charts, or you want short sessions on your phone. Older adults interested in speed-of-processing training should look at BrainHQ.
Do both if you have the time. Our 10-minute daily brain workout shows one way to rotate puzzles and apps through the week.
If you or a family member has MCI, the 2022 trial is a reason not to assume an app beats a familiar puzzle. It is one small study, though, and a doctor is the right person to talk to about a care plan.
Either way, pair it with physical activity, sleep, and social connection. See our guides on improving memory and healthy cognitive aging.
If you choose an app
BrainHQ is our top pick, with the deepest research record. It offers Double Decision, which BrainHQ says is an updated version of the speed training used in the ACTIVE trial. The trade-offs are a slow pace and a dated interface. $96 a year.
Elevate practices everyday skills like reading, writing, and math, though evidence that the gains carry over outside the app is limited. $39.99 a year.
For the full comparison, see our best brain training apps roundup.
Zenelia is a brain training app made by Mindkindly’s owner, so we don’t rank or score it here. It costs $6.99 a month or $44.99 a year, has a 7-day free trial, and is iPhone only for now; read more about Zenelia or see it on the App Store.
Common questions
Do crosswords prevent dementia? No one can promise that. Crossword use has been linked with a later start of memory decline in observational research, and one small trial in people with MCI favored crosswords over games, but the National Institute on Aging says no drugs or lifestyle approaches can currently prevent Alzheimer’s or a related dementia.
Are sudoku and word games like Wordle the same as crosswords? Not quite. Sudoku uses number logic rather than vocabulary, and short word games use a narrower slice of word skills. The head-to-head trials above tested crosswords specifically.
Paper or digital crosswords? Both trials used on-screen crosswords. Choose whichever format you will keep using.
How often should I do them? There is no proven dose. The 2022 trial used four 30-minute sessions a week, and the 2015 trial about 15 minutes at least five days a week.
The bottom line
Crosswords draw mainly on vocabulary and knowledge, cost little, and are easy to share. Apps cover a wider range of skills with difficulty that adapts, and BrainHQ offers an updated version of the speed training studied in the ACTIVE trial. The head-to-head research is split and limited, so neither is the clear winner. Pick the one you will actually do, keep it challenging, and treat it as one part of a mentally and physically active life.
Sources
- Computerized games versus crosswords training in mild cognitive impairment (Devanand et al., NEJM Evidence, 2022)
- Enhancing cognitive abilities with comprehensive training: a large, online, randomized, active-controlled trial (Hardy et al., PLOS ONE, 2015)
- Predictors of crossword puzzle proficiency and moderators of age-cognition relations (Hambrick, Salthouse and Meinz, Journal of Experimental Psychology: General, 1999)
- Association of crossword puzzle participation with memory decline in persons who develop dementia (Pillai et al., Journal of the International Neuropsychological Society, 2011)
- The impact of sustained engagement on cognitive function in older adults: the Synapse Project (Park et al., Psychological Science, 2014)
- Do “brain-training” programs work? (Simons et al., Psychological Science in the Public Interest, 2016)
- Ten-year effects of the ACTIVE cognitive training trial (Rebok et al., Journal of the American Geriatrics Society, 2014)
- Cognitive health and older adults (National Institute on Aging)
- Memory problems, forgetfulness, and aging (National Institute on Aging)
- BrainHQ subscription plans (BrainHQ Support)
- BrainHQ for libraries
- BrainHQ and Medicare offerings, full list for 2026 (BrainHQ Support)
- Elevate on the App Store