ADHD hyperfocus: why it shows up for some tasks and vanishes for others

Short answer

Hyperfocus is what happens when a task stops costing attention and starts generating it — you don't fight to concentrate, you fight to stop. It tends to appear when work is interesting, immediately rewarding, and has visible stakes, and it tends to vanish when work is abstract, delayed, or administratively dull. The same brain produces both.

Originally published on Founder And The City, Jane's newsletter about building in health tech. This version is written for the person trying to understand their own attention, rather than for the people managing them.

And if you can write twenty pages in a day but cannot open an email — that isn't laziness, and it isn't a character flaw. It's the most commonly misread feature of this whole picture, because the two things look like opposites and are in fact the same mechanism at two ends of its range. The wall you hit in front of a small task and the nine hours you lose to a big one are produced by one system, not two.

Jane, who has ADHD and runs a company, puts the practical conclusion this way: she spent years trying to fix the machine, and the thing that finally worked was changing the inputs.

What ADHD hyperfocus actually feels like

The descriptions people give are remarkably consistent. Time stops registering. You look up and four hours have gone. You forget to eat, or you eat without noticing. Someone speaks to you and you genuinely don't hear them. You come out of it slightly disoriented, often with a headache, and usually having produced a lot.

What it is not is chosen. People describe being pulled into it rather than deciding to enter it, which is exactly why "just hyperfocus on your actual work" is useless advice.

Two honest caveats before we go further.

Hyperfocus is a shakier scientific construct than the internet suggests. A 2021 review in Psychological Research found the term poorly operationalised across the literature, with studies measuring rather different things under the same word. That doesn't mean the experience isn't real — almost everyone with this brain recognises the description immediately — but it does mean you should be sceptical of anyone quoting precise figures about it.

It isn't exclusive to ADHD, either. Deep absorption in an engaging task is a normal human capacity. What differs is the range: how much more productive the good state is, how much harder the ordinary state is, and how little control you have over which one you get.

Why you can do the hard thing and not the easy one

Here is the reframe that makes the rest of this make sense.

Stop treating the wall as a fuel gauge reading empty. Treat it as a price.

Effortful, deliberate attention — the kind that doesn't happen by itself — appears to be allocated rather than simply depleted. A 2013 paper in Behavioral and Brain Sciences made the case that the feeling of mental effort works like an opportunity cost signal: your brain is continuously comparing what this task would cost against what that attention could be doing instead, and the sensation of "I can't face this" is the output of that comparison.

Which explains something that otherwise looks absurd. A forty-minute administrative task you are entirely capable of doing can be genuinely impossible, while a gnarly problem ten times harder is not. If that description lands hard, the wall in front of small tasks has its own article, and so does ADHD task paralysis, which is the more acute version of the same thing. The difficulty isn't the variable. The price is.

And you cannot argue with a price using willpower. Nobody stands in a shop shouting at a label. You change what's on offer.

There's a second mechanism stacked on top, and it's better evidenced. Delay discounting — how sharply a reward loses value the further away it is — is steeper in ADHD, confirmed across a 2016 meta-analysis. A deadline three weeks out has almost no present value in this brain. That isn't a metaphor and it isn't an attitude problem; it's a measured difference in how future rewards are weighted. Which is why "but it's due next month" fails as motivation, and why the same task becomes trivially doable the night before.

Put the two together and the pattern stops being mysterious. Work that is interesting, immediate and consequential is cheap. Work that is dull, abstract and distant is expensive to the point of impossible. The machine is working exactly as specified.

The best evidence that the room matters more than the person

This is the study worth knowing about, because it's the closest thing to a controlled experiment on the question.

In 2021, a Swedish group at Karlstad University took high-level youth athletes and had a trained psychotherapist score them on a standard adult ADHD symptom scale — twice. Once for how they function at school, once for how they function in their sport. Same young people. Same instrument. Same week.

The scores were high for school and low for sport.

Sit with that for a second. The symptoms did not travel with the person. They appeared in one context and receded in another, on a validated instrument, within days. Whatever the school assessment was measuring, it wasn't purely a property of the individual. It was a property of the individual plus the environment.

And this isn't simply "they try harder at the fun thing." Look at what sport structurally supplies: immediate feedback, a visible score, real stakes, varied and intense stimulation, and consequences that arrive in seconds rather than at the end of term. That is, almost item for item, the list of conditions under which an attention system that discounts the future steeply performs well. Nobody designed sport that way out of kindness — it was already like that, because that's what makes it worth watching.

Prevalence data points weakly in the same direction: a 2019 narrative review in the British Journal of Sports Medicine put ADHD at roughly 7–8% among student and elite athletes against something like 3–7% in the general population. Overlapping ranges, and a narrative review rather than a meta-analysis, so the honest reading is "somewhat over-represented" rather than anything dramatic.

The counterweight matters too, and it makes the point stronger. A 2023 scoping review of neurodiversity in elite sport found 23 studies, most concerned with ADHD as a concussion risk factor or with keeping medicated athletes compliant with anti-doping rules. A genuinely supportive environment is still described in that literature as future work. Sport did not get enlightened. It simply happens to be shaped correctly — and the trait stopped presenting as a deficit anyway.

That's the actionable version: the same person, in a differently shaped environment, measures differently.

So how do you actually trigger hyperfocus?

You mostly can't, on demand, for a chosen task. What you can do is make it much more likely to land somewhere useful. Four things do most of the work.

Keep many things open at once, not a queue. A queue has an order, and the order is exactly the part that doesn't function. What works better is a fan: several genuinely valuable problems all live at the same time, none of them designated as next. You look at what's laid out and notice which one is pulling — and that pull is rarely subtle. This trades predictability on any single task for far more total time spent in the cheap state.

Curate what's available. Hyperfocus is a mechanism, not a judgement. It will land on a Wikipedia rabbit hole about ball bearings as happily as on your real problem, and six hours reorganising a database nobody needs is not a superpower. But that proves the input was poor, not the mechanism. If your environment offers endless cheap stimulation and no accessible meaningful work, it will faithfully amplify the cheap stimulation. The real work happens before the work: maintaining an actual inventory of interesting, valuable, accessible problems rather than a backlog of administrative debris.

Protect the state once you're in it. Switching cost is the hidden expense. A status update, a Slack thread, a routine meeting pulled into the middle of a productive session doesn't cost fifteen minutes — it costs the configuration, which took much longer to reach than it takes to lose. Constant task switching has a measurable cost for everyone; here it's simply larger.

Build slack around it. This is the part people skip. If you run the expensive end of the system hard, you need genuinely quiet days around it, deliberately kept clear. Not as a reward and not as recovery-with-guilt — as a structural requirement. The alternative isn't more output; it's the kind of flat exhaustion that doesn't lift at the weekend — burnout with better vocabulary.

The thing that actually costs you: the guilt loop

Worth naming separately, because it's the most expensive part and the least visible.

You can't start the task, so you feel guilty. But the guilt isn't free — it runs on the same budget as the work. Deliberate, effortful, self-directed processing is precisely what the task needed, and self-criticism consumes it. So you have less. So you do less. So you feel worse.

Guilt isn't a motivator. It's an expense, billed to the same account as the work.

Which is the only reason the next sentence isn't a wellness platitude: sometimes the system needs to run in passive mode, and rest taken with guilt attached is the worst of both — you pay the tax and get none of the benefit. Doing nothing on purpose, without turning it into either a moral failure or another optimisation project, is a functional requirement rather than a treat.

What doesn't work, and why people keep trying it

Most productivity advice was written by people in the narrow range, for people in the narrow range. Two pages a day even if they're terrible. Always commit to a deadline. Eat the frog. Inside that range the advice is genuinely excellent.

Run it on a wide-range system and it produces a specific, recognisable failure: a few good weeks, then collapse, then the conclusion that you are the problem. On schedule, roughly every three weeks, for years.

The same applies at the organisational level. When someone's output looks lumpy, the reasonable interventions all point one way — fewer projects, no switching, tighter dates, more check-ins, visible daily progress. Each is defensible on its own terms, and the immediate result is reassuring: the board looks clean, control is restored.

The real result shows up a quarter later, as less deep work, fewer breakthroughs, more cognitive budget spent on self-coercion, and eventually either falling output or someone quietly leaving. By then the feedback is too delayed to be connected to its cause, so the conclusion drawn is that the person needed more control.

The cheap fix is unglamorous: widen the measurement window. Ask what valuable output happened over a quarter, not only whether the assigned ticket landed by Thursday.

This isn't a permission slip

The obvious objection: so people with ADHD should do whatever they feel like and let everyone else absorb the consequences?

No. Freedom without design isn't an operating model.

If the environment provides several genuinely valuable problems, flexibility in sequencing, long uninterrupted blocks, and a longer measurement horizon, then the other half of the deal has to be real too: honest communication about uncertainty up front rather than after the miss, responsibility for the value of what you choose rather than just the pleasure of choosing, visibility into what's moving and what's stalled, and safeguards so nothing critical depends on you alone.

And crucially, actual delegation or accommodation of the tasks that cause disproportionate paralysis — not a promise to fix yourself. Jane's version is blunt: she never solved the email problem, and someone else books her appointments. What makes that legitimate rather than a cop-out is that she stopped pretending she'd fix it and paid for a solution instead.

Some jobs genuinely are predictable execution, and that is the product. If you're in the critical path for a synchronised daily delivery and can't reliably show up, that's a routing error rather than a moral failure. Get out of the path. But a great deal of valuable work — research, strategy, invention, early product, crisis, finding the hidden constraint — has no honest deadline anyway, and that's the natural habitat for this kind of attention.

When to see a doctor about ADHD symptoms — and what to bring

Nothing above is a substitute for assessment, and self-diagnosis from an article is a poor idea in either direction — both "I definitely have this" and "so I don't need help" are conclusions this page can't support.

Consider seeking assessment if difficulty with attention, organisation, or starting tasks has been present since childhood, shows up across more than one area of your life, and is causing real problems at work, in study, or in relationships. Adults often present with the organisational and emotional side rather than the stereotyped hyperactivity, and it's routinely missed in women.

Speak to a doctor sooner if you're experiencing persistent low mood, anxiety that doesn't settle, or thoughts of harming yourself. Those need attention in their own right and shouldn't wait behind an ADHD question.

What to bring. Concrete examples rather than adjectives: specific tasks you couldn't start and what happened, school reports if you can find them, and — often the most useful thing — an account from someone who knew you as a child. If you've tried strategies, bring what you tried and what happened, because "I've tried X, Y and Z and here's the result" is a very different conversation from "I'm disorganised."

What to ask. Whether a full assessment is indicated and what the referral route is where you live; whether anything else could be producing the same picture, since sleep disorders, thyroid problems, iron deficiency, depression and anxiety can all look similar; and if you're already diagnosed, whether your current management still fits the life you actually have.

If you're waved off with "everyone's a bit like that" — the distinction that matters clinically is impairment across settings and duration, so bring the evidence of both. Ask for the reasoning to be recorded in your notes. That request alone tends to change the conversation.

Start understanding your body

Learn what affects your energy, stress, sleep, and daily state. Get the app.

Explore your data

How Welltory fits — and what it can't do

Welltory does not detect, diagnose, or measure ADHD, and no wearable can. What it can do is make the cost of a working pattern visible, which is a smaller claim and a more useful one.

What a long hyperfocus session actually costs. People routinely describe a nine-hour session as free because it felt effortless. The stress and recovery data usually disagree — and seeing that is often the first argument that lands for someone who has spent years being told to just work harder.

Whether your quiet days are actually quiet. The slack days only do their job if the body treats them as slack. If a "rest day" shows the same load as a working one, the schedule isn't doing what you designed it to do.

Sleep, honestly measured. Sleep debt worsens executive function in everybody, and the working pattern described in this article is unusually good at producing it. This is the single most modifiable thing on the list.

Whether a change actually helped. If you restructure your week — fewer meetings, longer blocks, protected quiet days — it's worth having something other than memory to judge it by, because memory is heavily coloured by whatever happened yesterday.

What Welltory can't tell you is which task to work on, whether your attention pattern warrants assessment, or anything at all about your diagnosis. Those questions live with you and your clinician.

Read more from Jane

Jane writes about building in health tech, systems, and the human side of running a company in Founder And The City. The original essay goes considerably further than this version — into Goldratt's bottleneck theory, Taleb's Mediocristan and Extremistan, why venture capitalists think in power laws right up until they open a headcount spreadsheet, and the exact deal she runs with her own team. The discussion of it is on LinkedIn.

Discounts for blog readers: up to 36% off

See what affects your energy, stress, sleep, and daily state with Welltory

Explore your data