← All articles

QZBrain Journal

Brain Exercises Worth Doing — Each One Graded Honestly

July 2, 2026·10 min read

You want to keep your mind sharp, and you have probably heard a hundred confident, conflicting claims about how. Here is the honest version, up front.

The most powerful things you can do for your brain are not puzzles or apps. The strongest evidence is for moving your body and sleeping well, and — on more associational grounds — staying socially and mentally engaged. Everything else on this page is a complement to those, never a substitute.

That does not mean puzzles and brain training are pointless. They are fun, they give a genuine workout to logic, attention, and memory, and the practice is real. What the evidence does not support is the promise that any of them will make you broadly smarter or shield you from decline.

Below is a graded rundown. Each item gets a plain what it does and what it does not do, plus a link to a deeper post if you want the full story.

How to read these grades

Two words do most of the work in this whole field: near transfer and far transfer.

Near transfer means you get better at the thing you practice and at tasks very close to it. Far transfer means the benefit spills over to unrelated abilities, like your general intelligence, your work, or your everyday memory. Near transfer is well established. Far transfer is the part that keeps failing to show up.

In the largest study of its kind, Owen and colleagues (2010) trained more than 11,000 people online for six weeks. People improved on the exact tasks they practiced, but those gains did not carry over to untrained tasks. A broad field review by Simons and colleagues (2016) reached the same shape of conclusion: robust gains on trained tasks, modest near transfer, and little to no far transfer.

So read each grade below through that lens. Worth doing does not have to mean rewires your brain. Most of these are worth doing because they are enjoyable, focused, and good for the specific skills they build.

This is not a counsel of despair; it is a calibration. Once you stop expecting a crossword to raise your general intelligence, you are free to enjoy it for what it genuinely delivers: focus, satisfaction, and a small daily win. That is lowered claims, not lowered value.

Do the puzzle because you enjoy the puzzle, not because a marketing page promised it would rebuild your brain.

Puzzles: sudoku, word search, nonograms, crosswords

Grade: genuinely worth doing, with realistic expectations. A good puzzle is a real logic and attention workout and a calm, absorbing ritual.

What they do: each puzzle drills a specific mental skill in a satisfying loop.

• Sudoku rewards pure logical deduction and patient scanning of a grid.
• Word search trains visual search and sustained attention as your eyes sweep for a target.
• Nonograms, also called Picross, blend logic and spatial reasoning as you deduce a picture from number clues.
• Crosswords lean on vocabulary and long-term memory retrieval.

There is also a quieter benefit that rarely makes the marketing copy: a puzzle is a contained, low-stakes way to practice paying attention. In a day full of pings and half-finished tasks, ten minutes of single-minded focus on one grid is its own reward.

What they do not do: there is no strong evidence that solving sudoku or word searches raises your general intelligence or protects you broadly. What you reliably get is better at that puzzle. That is a perfectly good reason to play.

Play free puzzles in your browser → Sudoku, word search, and nonograms, free and with no account.

If you would rather improve than plateau, each of these rewards a little technique.

Sudoku techniques that actually help → Turn guessing into real deduction with a handful of scanning methods.

Word search for adults → Why the humble word search is a better attention drill than it looks.

A beginner's guide to nonograms → How the number clues work, and how to start solving without guessing.

Learning an instrument or a language

Grade: among the better use-it-or-lose-it bets, not because it is proven to boost intelligence, but because it is rich, novel, effortful, and often social.

What it does: sustained learning of something genuinely new pulls in memory, attention, motor control, and hearing all at once, and it keeps doing so for years, because there is always a next level. A language adds a social dimension, conversation, culture, connection, that a solo puzzle cannot.

Why single this one out at all, then? Because the ingredients that seem to matter most for staying mentally engaged, novelty, sustained effort, genuine difficulty, and social contact, tend to show up here all at once. That combination is hard to beat, even when no single study can promise you a specific outcome.

What it does not do: it is not a magic bullet, and this is where it pays to be careful. The evidence that any single hobby prevents cognitive decline is thin, and studies of engaged people struggle to separate cause from the fact that healthier, more active people tend to take up more hobbies in the first place. Treat it as one of the more promising ways to stay engaged, not as a guarantee.

Dual n-back and working-memory training

Grade: a demanding, legitimate workout for working memory, with a far-transfer claim that remains unproven.

Dual n-back is the poster child here. It is hard, and it clearly makes you better at itself. The controversy is whether it does anything more than that.

The excitement started with Jaeggi and colleagues (2008), who trained 70 people on dual n-back across several dosing groups and reported gains in fluid intelligence, with more training days linked to bigger apparent gains. But a carefully placebo-controlled study by Redick and colleagues (2013) ran roughly 20 sessions and found no positive transfer to fluid intelligence or other abilities. A meta-analysis by Au and colleagues (2015) landed in between, estimating a small n-back-to-intelligence effect of around g = 0.24, real, but modest and much debated.

The broader working-memory literature tells the same story. Melby-Lervag and Hulme (2013) found that working-memory training produces short-term gains on working-memory tasks but no convincing spread to other skills.

Dual n-back: does it really work? → The full, honest read on the most over-hyped exercise in the field.

A calm guide to working-memory training → What working memory is, and how to train it without chasing a myth.

App-based brain training in general

Grade: worth doing for the habit, the task-specific gains, and the self-awareness of a trend, not for a promise to make you broadly smarter.

This is the category with the worst marketing history, so it deserves the clearest honesty. In 2016, the US Federal Trade Commission announced that Lumosity would pay two million dollars to settle charges of deceptive claims that its games improve performance at work and school and guard against age-related decline and dementia. In 2014, a consensus statement signed by more than 130 scientists warned that there was no compelling evidence brain games deliver broad cognitive benefit, and cautioned against exploiting older adults' fears.

So what is a brain-training app actually good for? Three real things: it builds a consistent, low-friction habit; it delivers genuine task-specific gains on what you practice; and it can show you a personal trend over time, which is quietly motivating.

Of those three, the self-tracking is the one people underrate. Not because a number on a chart is magic, but because watching an honest trend of your own speed and accuracy is what keeps you coming back, and consistency is where the small, real gains actually accumulate.

There is one much-discussed bright spot, and it needs its hedge every time. In the large ACTIVE trial (Ball and colleagues, 2002), training improved the specific abilities it targeted. A later follow-up (Edwards and colleagues, 2017) reported that the speed-of-processing training arm was associated with about a 29% lower risk of a dementia diagnosis over ten years. Read that carefully: it was a single training arm, the dementia outcome was based on claims and self-report rather than a clinical work-up, and associated with is not proof of prevention. It is an intriguing signal, not a promise.

This is general information, not medical advice. If you notice sudden, persistent, or worsening changes in your memory or thinking, see a qualified professional rather than reaching for an app.

Does brain training actually work? → The honest anchor for everything on this page; read it if you read nothing else.

Processing-speed training, explained → What speed training can and cannot do, with the ACTIVE caveats kept intact.

Aerobic exercise: the strongest-evidence lever

Grade: the best-supported thing on this entire list, and it is not a brain game at all.

If you change only one habit for your brain, make it this one. A 2018 meta-analysis by Northey and colleagues found that moderate-to-vigorous exercise, in sessions of roughly 45 to 60 minutes, improved cognition in adults over 50. That is a stronger and more consistent signal than anything the puzzle or app world can offer.

You do not need a gym. A brisk walk, a bike ride, dancing, swimming, anything that raises your heart rate and that you will actually repeat, counts. The best exercise for your brain is the one you keep doing.

Exercise and brain health → Why movement beats every app in the evidence, and roughly how much you need.

Sleep and social engagement: the foundation

Grade: foundational. These are not optional extras you bolt on after the puzzles. They are the ground everything else stands on.

Sleep is not downtime for your brain; it is active maintenance. Diekelmann and Born (2010) describe how sleep consolidates memory, moving what you learned during the day into more durable storage. Skimp on sleep and no amount of training will land properly.

Social engagement belongs here too. Rich conversation, shared activities, and staying connected keep your mind working in ways a solo screen cannot fully replace. The evidence here is more about association than proof, so hold it loosely, but as a low-risk, high-reward bet, staying connected is an easy yes.

Sleep and memory → How a good night of sleep does more for recall than a night of drills.

Where QZBrain fits

If you want a single mental model, stack it like this, top to bottom, most evidence to least.

• Foundation: sleep well, and stay socially and mentally engaged.
• Strongest lever: regular aerobic exercise.
• Promising and enjoyable: novel, effortful learning, like an instrument or a language.
• Fun complements: puzzles and brain-training apps, for the habit and the specific skills they build.

Notice what is not at the top: the apps and the puzzles. They earn a real place on the list, just never above sleep, movement, and connection. Anyone who tells you a game outranks those is selling something.

The good news is that you do not have to choose. A short daily puzzle can be the enjoyable complement that sits happily on top of a well-rested, active, connected life.

QZBrain is built to sit exactly where the evidence says a brain-training app belongs: as an enjoyable complement, honestly framed.

It is free, works offline, and needs no account. Focus mode gives you calm baseline training and records a personal NeuroIndex, a trend built from your own speed, accuracy, and difficulty over time. Watch it over weeks like a running pace, not as a verdict, and never as an IQ or a diagnosis.

If you take one thing from this page, let it be the hierarchy: sleep, move, and connect first, then enjoy the training for what it really is.

Open QZBrain: free, offline, no account → Start a calm two-minute session and watch your own trend, honestly.

Start with the honest overview → The clear-eyed answer to whether any of this actually works.

Frequently asked questions

What are the best brain exercises?

The strongest evidence points to aerobic exercise, good sleep, and staying socially and mentally engaged, not to any single puzzle or app. Beyond that, activities that are novel, effortful, and enjoyable, like learning an instrument or a language, or a puzzle you actually like, are reasonable bets. Just expect to get good at the specific thing you practice, rather than broadly smarter.

Do brain exercises actually work?

They reliably make you better at the exercise itself and closely related tasks, which is called near transfer, and it is real. What large studies like Owen and colleagues (2010) and reviews like Simons and colleagues (2016) do not find is far transfer, a spillover to general intelligence, work, or school. So they work for what they are, not as a shortcut to a sharper mind overall.

What helps your brain more than brain games?

Regular aerobic exercise has the best evidence; a 2018 meta-analysis by Northey and colleagues found that moderate-to-vigorous exercise improved cognition in adults over 50. Consistent sleep, which actively consolidates memory (Diekelmann and Born, 2010), and staying socially engaged are also foundational. Treat games and apps as an enjoyable complement to those, never a replacement.

How often should you do brain exercises?

There is no magic dose, so pick a frequency you can actually keep; a few short sessions a week is fine, and consistency matters more than marathon sessions. The habit itself is one of the most durable benefits. If you enjoy it daily, great; if not, do not force it.

QZBrain is a general wellness and brain-training product for everyday cognitive exercise and entertainment. This article is general information, not medical advice, and is not intended to diagnose, treat, cure, or prevent any condition.