Self Lab

Evidence-based self-improvement. No mechanism, no article.

Esc or click outside to close

Attention & Dopamine · Mechanism

The 25-Minute Myth: Your Pomodoro Timer Isn't Science

Breaks reliably lift your mood and cut fatigue, but a meta-analysis of 2,335 people found no significant gain in performance. Here's what 25 minutes really is.

Alex by Alex July 29, 2026 17 min read

Open any productivity feed and you’ll hear the same claim: your brain focuses for about 25 minutes, so work in 25-minute blocks. Tidy. Also not where the number came from.

Francesco Cirillo built the Pomodoro Technique in the late 1980s as a university student, timing himself with a tomato-shaped kitchen timer (EBSCO Research Starters). He began with stretches of roughly ten minutes and lengthened them as his focus improved. No source, including Cirillo’s own, presents 25 minutes as a research finding. The official Pomodoro site doesn’t even try to explain the number.

So what does the evidence actually support? Something quieter, and more useful than a timer. If you already know that motivation is a mood, not a fuel tank, you’ll recognize where this lands.

The Bottom Line

  • The 25 minutes is Cirillo’s kitchen timer, not a study. He started nearer ten.
  • The study people cite for it used “breaks” of about two seconds inside a 40-minute lab task (Ariga & Lleras, 2011).
  • Breaks do lift energy and cut fatigue. Pooled across 2,335 people, their effect on performance wasn’t statistically significant (Albulescu et al., 2022).
  • What holds up is scheduling breaks in advance instead of deciding mid-session (Biwer et al., 2023).
  • No trial has ever tested 25 minutes against another length in real work. So pick a number you’ll actually keep.

In this guide:

Where does the 25 minutes actually come from?

From one student and one kitchen timer. Cirillo developed the technique in the late 1980s while at university, using a tomato-shaped timer, which is where the name comes from (EBSCO Research Starters). He began with roughly ten-minute stretches and lengthened them as his concentration improved.

That’s the whole origin story. A person experimented on himself and settled on a number that suited him.

Notice what’s missing. The official Pomodoro Technique site offers no explanation of why the interval is 25 minutes, and makes no scientific claim for it. That absence matters more than it looks. If a research basis existed, the people who own the method would be the first to name it.

You’ll also find blogs claiming 25 was chosen because kitchen timers only went up to 30 minutes. There’s no primary source for that either. It’s a story about a story.

None of this makes the technique useless. Cirillo never promised a neuroscience finding, and self-experimentation is a perfectly reasonable way to build a personal work rhythm. The problem starts later, when the internet promoted one person’s preference into a fact about human attention. Nobody checked. The number just kept getting repeated until it sounded like it had a lab behind it.

So where did the “attention span” part come from? A different mess entirely.

Can your brain really only focus for 25 minutes?

There’s no primary data behind that claim. A 2016 critical review in Advances in Physiology Education went looking for the evidence and concluded that “the available primary data do not support the concept of a 10- to 15-min attention limit” (Bradbury, 2016). The article is labelled a personal view, meaning it’s a review of the literature rather than a new experiment.

Here’s the part worth sitting with. Bradbury traced the famous 10 to 15 minute figure back through decades of citations to a single 1978 paper by Hartley and Davies. That paper wasn’t about attention span. In Bradbury’s words, “the subject of the manuscript is in fact ‘note taking.’” And it was itself a review, not a primary data source.

A citation chain, repeated long enough, started looking like a measurement.

The same paper dismantles the “8-second human attention span, one second less than a goldfish” line, a Microsoft-commissioned figure reported in Time in 2015. Nobody knows how to measure a goldfish’s attention span, and the 8 seconds actually measured how long people stay on a web page. Bradbury’s most consistent actual finding: the biggest source of variation in student attention was the teacher, not the format of the lesson.

The honest move here isn’t to replace 25 with a better number. It’s to stop believing the category. Your attention isn’t a battery with a published runtime. It swings with the task, the sleep you got, the room, and how interesting the thing in front of you is.

But doesn’t attention really decay over time?

On some tasks, yes, and this is the one real finding underneath the myth. In 1948, Norman Mackworth had people watch an unmarked clock face for two hours and report the occasional double-jump of the hand (Mackworth, 1948). The setup mimicked British radar operators on wartime watch. Detection accuracy dropped noticeably within the first half hour.

That’s genuine. It’s also a two-hour monotonous signal-detection watch, which is close to the opposite of self-directed work you chose and understand. Borrowing its timing and applying it to your essay draft is a stretch nobody has justified.

What did the study everyone cites actually test?

Two-second task switches. Not breaks. The paper productivity writers reach for is Ariga and Lleras (2011) in Cognition, and almost nobody who cites it has read the method section (Ariga & Lleras, 2011).

Here’s what happened. 84 students sat through a 40-minute visual vigilance task: 1,200 trials, split into four uninterrupted 10-minute blocks, pressing a key when a line looked shortened. Targets appeared on 10% of trials.

The “switch” group got digit-recall probes after trials 600, 900 and 1,200, and only the first two fell inside the vigil. Each probe flashed for about a sixth of a second, with under two seconds to respond. Add it up and that group got roughly four seconds of interruption across 40 minutes. Nobody rested. Nobody stood up. Nobody made coffee.

The result was real and interesting. The switch group’s detection accuracy stayed flat across all four blocks. Every other group’s declined significantly. There was no speed versus accuracy trade-off, and memory scores were basically identical between groups.

But look at what that supports. Briefly reactivating your task goal stops attention from decaying on a lab vigilance task. It says nothing about rest. It never tested a 25-minute interval, and it never tested a 5-minute break. The authors’ only practical suggestion, and it came in the press release rather than the paper, was to impose brief breaks on yourself during long tasks like studying for a final exam.

Straight talk: the study most often used to prove you need a 5-minute break every 25 minutes tested neither the 5 minutes nor the 25.

Do breaks actually make you produce more?

Breaks change how you feel more reliably than what you finish. A 2022 systematic review and meta-analysis in PLOS ONE pooled 22 independent samples and 2,335 people to test micro-breaks, meaning breaks of ten minutes or less (Albulescu et al., 2022).

The well-being results were clear. Vigor, roughly “feeling energised”, improved with an effect size of d = 0.36. Fatigue dropped with d = 0.35. Both were statistically solid. Treat d as a size label: bigger number, bigger difference.

Performance was a different story. The pooled effect was d = 0.16 and it wasn’t statistically significant, with a range running from slightly negative to modestly positive. In plain terms, the data can’t tell that effect apart from zero.

Then it gets more awkward. When the authors split by task type, cognitively demanding tasks came out slightly negative, at d = -0.09, though that one is no more distinguishable from zero than the overall figure. Creative and clerical tasks did better. And longer breaks, not shorter ones, predicted better performance.

Breaks help how you feel, not what you finish Albulescu et al., 2022, pooled 22 samples and 2,335 people. Vigor improved with d = 0.36 and fatigue fell with d = 0.35, both statistically significant. Overall performance was d = 0.16, not significant, with a confidence interval from minus 0.04 to 0.37. On cognitively demanding tasks the pooled effect was minus 0.09, which is also not significant. Breaks help how you feel, not what you finish Pooled effect sizes (d) across 22 samples, 2,335 people Vigor (feeling energised) 0.36 Less fatigue 0.35 Performance (all tasks) 0.16 not significant Performance (thinking-heavy) -0.09 not significant -0.1 0 0.2 0.4 0.6 The thin line marks the range around the performance estimate. It crosses zero. Source: Albulescu et al. (2022), PLOS ONE
Micro-breaks moved mood reliably. Output, not so much. Source: Albulescu et al. (2022).

The authors are upfront about limits: modest numbers of studies per outcome, self-reported measures, and pre-COVID data. Still, the shape is hard to miss. The five-minute break is a mood intervention with evidence behind it and a performance intervention without one.

Does the Pomodoro technique work at all?

It works, mostly for reasons the marketing never mentions. In 2023, researchers had 87 students study for a real day under three conditions (Biwer et al., 2023). One group chose their own breaks. One took 6 minutes off after every 24 minutes. One took 3 minutes off after every 12.

The students choosing for themselves came off worst. They ran longer sessions, took longer breaks, and reported more fatigue and distraction plus less concentration and motivation. Both scheduled groups fared better on how the session felt.

Now the part that gets left out. Mental effort and tasks completed didn’t differ between any of the three groups. And the benefit is reported for scheduled breaks as a class. The paper gives no evidence that 24 minutes beat 12.

Two years later the same research group ran the real 25/5 schedule with 94 students, against self-chosen breaks and against “flowtime”, where you break when you choose and scale the break to how long you worked (Smits, Wenzel & de Bruin, 2025). Under Pomodoro, fatigue climbed faster and motivation fell faster than with self-chosen breaks, though the groups ended the session at the same overall levels of fatigue and motivation. Nothing separated the groups on productivity, task completion or flow.

So the honest summary isn’t “Pomodoro is great” or “Pomodoro is junk”. It’s that scheduling helps the experience, the specific interval is unproven, and the output gains people expect aren’t in the data.

The study that disagrees, and why we’re still telling you

A 2025 scoping review in BMC Medical Education covered 32 studies and 5,270 participants, and concluded that time-structured Pomodoro interventions consistently improved focus, cut mental fatigue and boosted sustained performance (Ogut, 2025). That’s the opposite headline. Leaving it out would be the same overclaiming we’re criticising.

Four things keep it from settling the question. It’s a scoping review, not a meta-analysis, so there’s no pooled effect size and no formal risk-of-bias assessment, only an adapted JBI quality appraisal. Only 3 of the 32 studies were randomised controlled trials. Its literature search closed in May 2023, so it couldn’t include the 2025 study that found fatigue rising faster. And its own structured comparisons were 24/6 and 12/3, which again is not a defence of 25.

What’s actually doing the work, if not the number?

Deciding before you start. Call it pre-commitment: you set the rule in advance, so the tired version of you in hour two doesn’t get a vote. That’s our reading of what separated the groups in the 2023 study, not a label the authors used. They framed it as effort regulation. Either way, the self-regulating students had to judge, in the moment, whether to push on. They judged badly.

The timer isn’t reading your brain. It’s removing a decision from someone with a bad track record on that decision, which is you at minute 40 with a half-written paragraph and a phone nearby.

That reframe changes what you optimise. If the number were doing the work, you’d hunt for the perfect interval. If pre-commitment is doing the work, the winning move is picking any sane block length and writing it down before you begin. It’s the same logic behind starting smaller than feels sensible: remove the in-the-moment negotiation and the plan survives contact with a bad day.

There’s an irony here that nobody markets. Pomodoro is sold as a focus tool, and the strongest evidence for it is about mood and session structure. Meanwhile the thing people credit, the magic 25, has never once been tested head to head against another interval in real knowledge work. Not in 2011, not in 2023, not in 2025.

If quitting mid-session is your actual problem, the break schedule isn’t the fix. Wanting to escape a task is usually an emotion problem, not a time problem.

What about “23 minutes to refocus” and the 52/17 rule?

Both are weaker than they sound. The famous “it takes 23 minutes and 15 seconds to get back on task” line has no paper behind it. It comes from a 2006 Gallup Business Journal interview with researcher Gloria Mark (Gallup, 2006).

The published study is real and much more interesting. Mark and colleagues shadowed 24 information workers across a 13-month, two-phase study with more than 700 hours of observation (Mark, González & Harris, 2005). People spent an average of 11 minutes 4 seconds on a working sphere before switching or being interrupted. 57.1% of those segments got interrupted. When interrupted work was resumed the same day, which happened 77.2% of the time, resuming took an average of 25 minutes 26 seconds.

Now the number nobody quotes: the standard deviation on that average was nearly 55 minutes. When the spread is twice the average, the average tells you almost nothing. And in the interview itself, Mark called the recovery time “not so long”. The productivity internet built a doom statistic out of a researcher shrugging.

The number you’ve heardWhere it really comes fromWhat it can honestly support
”Focus 25, break 5”Cirillo’s kitchen timer, late 1980sA workable schedule, not a finding
”You can only focus 10 to 15 minutes”A 1978 review about note-taking, retold for decades (Bradbury, 2016)Nothing about attention limits
”23 minutes 15 seconds to refocus”A 2006 Gallup interview. The published figure is 25 min 26 sec, with a 55-minute spreadInterruptions cost time, wildly variable time
”Work 52, rest 17”A time-tracking company’s blog post about its own usersWhat that app’s heaviest users happened to do

That last row deserves a note. The 52/17 rule is often presented as the evidence-based upgrade to Pomodoro. It isn’t a study at all. It’s DeskTime, a time-tracking company, describing the pattern of its own top 10% of users, published on the company blog. No control group, no comparison, and a product to sell.

How do you build a break schedule that survives a real day?

Pick a length you’ll keep, then take the choice away from yourself. Since no trial has compared 25 minutes to any other interval in real work, chasing the “correct” number is wasted effort. What did show up in the meta-analysis is that longer breaks predicted better performance, not shorter ones (Albulescu et al., 2022).

Step 1: Choose the block from your task, not from a blog

Fiddly admin and email? Short blocks are fine, and clerical tasks showed the largest break benefit in the 2022 meta-analysis, from just two studies. Writing, coding, studying, anything that takes ten minutes just to load into your head? Go longer. Those were the tasks where micro-breaks came out slightly negative, though not distinguishably so.

Step 2: Write the schedule down before you begin

This is the whole mechanism in one line. Decide the block and the break before your first minute of work, out loud or on paper. The 2023 study’s self-regulating group made that call mid-session and ended up more tired and less focused than either scheduled group.

Step 3: Make the break a real stop

Stand, walk, look out a window. Honest caveat: the meta-analysis didn’t analyse what people did during breaks, so treat this as a sensible preference rather than a proven rule. What it did test was breaks of ten minutes or less, which is the range most people can actually take.

Step 4: Don’t cut a session that’s genuinely flowing

The 2025 study included a “flowtime” condition, where students broke when they chose and scaled the break to the work behind it. It finished as much as the strict 25/5 group. Interrupting real momentum to obey a tomato is a rule serving itself.

Step 5: Judge the day by what you finished

Not by pomodoros counted. Across all three groups in the 2023 study, tasks completed didn’t differ. A tally of timer cycles measures obedience, not output. If your phone is the thing ending your sessions early, that’s a different loop with a different fix.

FAQ

Is the Pomodoro technique scientifically proven?

No, not as a specific 25/5 schedule. Scheduled breaks beat self-chosen ones on how a study session feels, in an 87-student study (Biwer et al., 2023). No trial has tested 25 minutes against another interval in real work, and task completion didn’t differ between groups.

Why 25 minutes specifically?

Because that’s what worked for Francesco Cirillo. He developed the method as a university student in the late 1980s with a tomato-shaped kitchen timer, starting around ten minutes and lengthening the interval over time (EBSCO Research Starters). The official site offers no scientific explanation for the number.

Is 52/17 better than 25/5?

There’s no evidence either way. The 52/17 figure comes from a time-tracking company analysing its own users’ logs and publishing the result on its blog. No control group, no comparison condition. Treat it as one company’s usage data, not as a rival to a research finding.

How long can a person actually focus?

There’s no fixed number, and the popular ones don’t survive checking. The 10 to 15 minute claim traces back to a 1978 paper about note-taking (Bradbury, 2016). Attention shifts with the task, your sleep and your interest. Build for your worst day instead, using two-minute starters.

Should I stop using Pomodoro?

Only if it isn’t helping you. Deciding the schedule in advance is the part with support behind it, so if the timer gets you started and keeps sessions from sprawling, keep it. Just drop the belief that 25 is a number your brain agreed to.

The Bottom Line

The timer isn’t reading your attention span. There is no published attention span for it to read. What the evidence supports is far less exciting and far more usable: deciding your breaks before you start beats deciding them at minute 40, when you’re tired and negotiating with yourself. Breaks will reliably make the session feel better. They probably won’t make you finish more, and on hard thinking work the effect may lean the wrong way. So stop auditioning intervals. Pick one you can repeat on an ordinary Tuesday, write it down, and judge the day by what got finished.

One small action today: before your next work session, say the schedule out loud, block length and break length, then set the timer. Any sane number. The point is that you chose it before you were tired, not after.

Alex is the voice of Self Lab: practical psychology for people who are done with motivational fluff.

Sources