A pomodoro timer for programmers who don’t work in 25-minute pieces

PomoFit is a browser-based pomodoro timer that puts a short exercise into every break. The interval is yours to set. This page is mostly about why the default is wrong for programmers and what to set instead.

25/5 was a kitchen timer, not a finding

Francesco Cirillo picked 25 minutes in the late 1980s using a tomato-shaped kitchen timer. That is the entire origin story. No lab derived it. Anyone selling you “the scientifically optimal focus interval” is selling you a countdown.

There are roughly four peer-reviewed studies on the Pomodoro technique in existence, and they are not flattering. Smits et al. 2025 (Behavioral Sciences, 94 students, two-hour randomised session) compared 25/5 against Flowtime and against letting people take breaks whenever they wanted. No significant difference in productivity, task completion or flow. The Pomodoro group got tired faster (p=0.025) and lost motivation faster than the self-regulated group. Biwer et al. 2023 found systematic breaks helped mood, and made no significant difference to mental effort or task completion.

We sell a pomodoro timer. We would rather you heard that from us.

What survives the scrutiny is narrower and more useful: structured breaks reliably restore energy and cut fatigue. That is the claim this product stands on.

The interval problem is a context-reload problem

The break is not the expensive part. The reload is.

Rebuilding the mental model after an interruption costs minutes, and a 25-minute timer often fires right after you finished paying that cost. Fixed short intervals tax exactly the work that takes longest to get back into.

The meta-analysis does not settle focus-block length, and nothing does. What Albulescu et al. 2022 (PLOS ONE, 22 samples, N=2,335) does show is that within breaks, longer ones did more for performance than shorter ones, on an overall performance effect that was not significant. Short and rigid has nothing arguing for it.

So set the timer to your work, not to a 1980s kitchen appliance:

• Deep debugging, new architecture, unfamiliar codebase. 90 minutes of focus, 15-minute break. One reload per session, not three.

• Normal feature work. 50 minutes of focus, 10-minute break. The default worth starting from.

• PR review, ticket triage, inbox, chores. 25 minutes of focus, 5-minute break. Shallow work is where classic pomodoros genuinely shine.

Start at 50/10. Move to 90/15 on the days you are somewhere hard.

“Stopping mid-flow to do push-ups sounds terrible”

It does. So don’t.

That objection is correct. A timer that fires blind into deep work is a bad timer. Two rules fix it.

Rule one: skip freely. Flow is rarer than the break. When it arrives, hit skip. PomoFit has no streaks to break, no chain to maintain, no confetti, no badge withheld. There is a number on the screen and nothing to click. Leaderboards exist as a Pro feature and you never have to look at them. A timer that punishes you for shipping is a timer you will uninstall in nine days.

Rule two: anchor the break to a moment your flow already ended. You have more of these than most desk workers, and they are free:

• A full build or a cold compile

• A CI pipeline you are watching for no reason

• The test suite on a large repo

• Waiting on a code review, or waiting on someone else’s answer in Slack

• Right after a merge, right after a deploy

• When you are stuck and the answer is not coming

That last one is the strongest. Ariga & Lleras 2011 (Cognition) found brief switches to an unrelated task roughly every 10 minutes prevented the decline that every control group showed across a 40-minute vigilance task. It was a vigilance task, not code, so treat it as an analogy. But “get up when you are stuck” is not a novel idea to anyone who has found the bug in the shower.

You are already standing there staring at a progress bar. The only change is what you do with the ninety seconds.

What the evidence supports, and what it does not buy you

Not “you will write better code”. Here is the real shape of it.

Albulescu et al. 2022 found micro-breaks of ten minutes or less significantly increased vigour (d=0.36) and reduced fatigue (d=0.35). The performance effect was d=0.16 and not statistically significant (p=.116). For cognitively demanding tasks, which is to say yours, it was d=−0.09. Nothing.

Reed & Ones 2006 (158 studies, 13,101 people) found acute aerobic exercise reliably raises energy and vigour, d=0.47, lasting at least 30 minutes, and low intensity worked consistently. You do not need to destroy yourself. Ten squats is a dose.

Diaz et al. 2026 (BJSM) had 11,484 US adults take five-minute movement breaks for 14 days. Fatigue and low mood fell, work engagement rose 4–7% and self-rated performance 1–3%. Uncontrolled and self-reported, so hold the numbers loosely. The direction is the part worth keeping: the breaks did not cost people output.

So the offer is energy at 4pm and a body that still works at 45. Not IQ points.

The part that is not about today

This is the real reason to bother.

Patterson et al. 2018 (European Journal of Epidemiology, 34 cohorts, 1.33 million people) found all-cause mortality risk starts rising above eight hours of sitting a day, and cardiovascular mortality above six. Count your day honestly. Commute-free desk work clears eight without trying.

Chen et al. 2018 (Physical Therapy, 27 RCTs) is the one that shaped the product. For desk-worker neck and shoulder pain, strengthening exercise worked, with moderate-quality evidence. Stretching was no better than not stretching. Ergonomic changes alone did not work, and neither did education. That is why PomoFit gives you air squats and push-ups instead of neck rolls, and why a new chair on its own is not a plan.

Ahmed et al. 2025 (BJSM meta-analysis, 11 trials, 414 inactive adults) found short exercise breaks significantly improved cardiorespiratory fitness with moderate certainty, at 91% adherence and 83% retention. It also found no significant effect on body composition, blood pressure or blood lipids. If you came for fat loss, this is not the tool.

And the permission slip: the US Physical Activity Guidelines (HHS, 2018) removed the old rule that activity had to come in bouts of at least ten minutes to count. Move more, sit less. Four minutes counts now.

A 10-minute break that is worth taking

Concrete, no equipment, fits between a build and a rebase:

1. 0:00–1:30 Stand up. Walk to the kitchen. Fill the glass.

2. 1:30–7:30 Three rounds: 10 air squats, 8 push-ups (knees or a desk edge is fine), 30-second wall sit.

3. 7:30–10:00 Walk back. Sit down. Read the failing test again with fresh eyes.

If you would rather not decide, run a ready-made plan. Full-body strength is the default for most developers. Energy-boost cardio is for the 3pm slump. Back-pain desk relief is for the lower back that started talking to you last spring. All twelve are at /plans.

Free tier, no account, nothing to install. PomoFit Pro is 4.99/month for personalised schedules, custom exercises, YouTube ambient audio and detailed stats. If you want more on the interval question, read why the Pomodoro technique stops working and what to do during a Pomodoro break. If you want the evidence with none of the sales copy, read developer fitness without a gym.