Does Brain Training Actually Work? What the Science Really Says

Does brain training work? Yes and no — and the distinction is the whole story. The research is consistent: you reliably get better at the specific tasks you practice, sometimes dramatically. What the evidence does not support is the bigger promise the industry once sold — that playing abstract games will raise your general intelligence, sharpen your memory across the board, or ward off cognitive decline. In short, brain training works on what you train, and not much beyond it.

That sounds like a letdown, but it's actually a useful map. Once you know that improvement is specific, you stop hoping a tile-matching game will make you smarter in general and start training the real skills you care about. Let's walk through what the science actually found, why the difference between "near" and "far" transfer matters, and how to train in a way that pays off.

What the big studies found

The most famous test of brain training was deliberately enormous. In 2010, neuroscientist Adrian Owen and colleagues published a study in Nature — run with the BBC's "Bang Goes the Theory" program — in which more than 11,000 people trained online for six weeks on games designed to improve reasoning, memory, attention, and planning. Participants did get better at the trained games. But when researchers tested whether that improvement carried over to broader, untrained cognitive abilities, they found essentially nothing. The authors concluded that there was no evidence the gains generalized to general cognitive function, even when training was substantial.

A few years later came the definitive sober second look. In 2016, a group of leading psychologists led by Daniel Simons published a sweeping review in Psychological Science in the Public Interest, evaluating the evidence behind commercial brain-training claims. Their verdict: the field had produced lots of studies showing improvement on trained tasks and on closely related tasks, but little credible evidence of "far transfer" to everyday cognition or real-world performance. Many positive results, they noted, came from studies with weak control conditions or small samples.

The pattern is remarkably consistent: practice a skill and you improve at that skill. Expect it to upgrade your whole mind, and the evidence quietly disappears.

It wasn't just academics raising an eyebrow. In January 2016, the U.S. Federal Trade Commission announced that Lumosity, one of the best-known brain-training brands, would pay $2 million to settle charges of deceptive advertising. The FTC said Lumosity had claimed its games could improve performance at work and school and stave off age-related decline, Alzheimer's, and other conditions — without the science to back it up. The settlement became a landmark moment, a formal acknowledgment that the marketing had outrun the evidence.

Near transfer vs. far transfer

The key concept hiding behind all of this is transfer — whether practicing one thing improves another. Researchers split it into two kinds:

• Near transfer: improvement spreads to closely related tasks. Practice a specific type of mental arithmetic and you'll likely get faster at similar arithmetic. This is well supported.
• Far transfer: improvement spreads to distant, unrelated abilities. Play a memory game and become a better problem-solver, a sharper driver, or a higher scorer on an IQ test. This is the prize brain-training ads promised — and it's the part the evidence keeps failing to deliver.

Here's the honest summary in one table:

Why is far transfer so stubborn? Because getting better at a task often means learning that task's specific tricks — its patterns, shortcuts, and quirks — rather than upgrading a general-purpose mental engine. Your brain is wonderfully adaptable, but it adapts to exactly what you ask of it. (We unpack the biology of that adaptability in neuroplasticity explained.)

So what actually moves the needle?

If you want broad cognitive benefits — better attention, memory, and processing across the board — the most evidence-backed intervention isn't a brain game at all. It's aerobic exercise. Meta-analyses of randomized controlled trials consistently find that regular cardio produces modest but real improvements in cognition, especially executive functions like working memory and cognitive flexibility, with the clearest effects in older adults. The mechanism is plausibly general — better blood flow, more of the growth factors that support neurons — which is exactly the kind of whole-brain upgrade games keep failing to provide.

Other well-supported habits matter too: good sleep, social connection, managing stress, and staying mentally engaged with genuinely challenging, novel activities. None of these is as flashy as a "make yourself smarter" app, but they're what the science actually endorses.

So where does that leave structured brain training? In a more honest, more useful place. If the lesson is you get better at what you practice, then the smart move is simple: practice the real skill, not a proxy for it.

How to train so it actually pays off

This is the principle we built Rusty around, and it falls directly out of the research:

1. Train the actual skill you want. If you want to do mental math faster, practice mental math. If you want to think and speak more clearly, practice reasoning and speaking out loud — not abstract shape-rotation that "supposedly" generalizes. Specificity is the feature.
2. Get real feedback. Repetition without feedback just rehearses your current level. Feedback is what converts reps into progress. (Rusty's AI-scored speaking, for example, grades the structure and clarity of a spoken answer and gives you one concrete thing to fix — that's near transfer working for you, because the skill you practice is the skill you want.)
3. Be consistent, not heroic. A focused five-minute brain workout routine done daily beats an occasional marathon. Skills consolidate through repeated, spaced practice.
4. Keep moving. Pair any mental practice with regular aerobic exercise — it's the closest thing to a general cognitive enhancer we have.

This is also why we're allergic to hype. As we argue in brain training in the age of AI, the goal isn't a magic IQ boost — it's keeping the specific human skills you rely on from going rusty while machines do more of your thinking.

Key takeaways

• Brain training works — narrowly. You reliably improve at what you practice (near transfer); broad "smarter overall" gains (far transfer) are weakly supported at best.
• The big evidence is sobering. Owen et al. (2010) found no transfer to general cognition across 11,000+ people; Simons et al. (2016) found little credible far-transfer evidence; the FTC fined Lumosity $2 million in 2016 for overclaiming.
• For broad benefits, exercise wins. Aerobic activity has the strongest evidence for general cognitive improvement.
• Train the real thing. Practice the exact skill you care about, get feedback, and stay consistent. That's how training actually pays off.

The honest verdict is also the empowering one. Brain training isn't a shortcut to genius, but it is a reliable way to get better at the things you deliberately practice. So choose those things well — and train them on purpose.

Want to train the real skills instead of a proxy? Download Rusty free on the App Store and run your first five-minute circuit today.