Box Breathing for Focus: A Rhythm Guide

Two Minutes Between Work Blocks

One work block has ended, the next starts in a couple of minutes, and the gap needs filling with something other than another browser tab. A common suggestion for that slot is box breathing: in for four, hold for four, out for four, hold for four, repeated until the time is up, often alongside the pomodoro convention.

The break side of that pairing rests on its own body of work, and it is more equivocal than the advice usually admits.

A 2022 meta-analysis of micro-breaks pooled twenty-two independent samples and about 2,300 participants. Short breaks produced small but statistically reliable gains in vigour and reductions in fatigue, and a non-significant effect on overall performance. Breaking that down by task type, the performance benefit appeared only for tasks with lower cognitive demand, and a meta-regression found that longer breaks helped performance more — the authors conclude that recovering from highly depleting work may need more than ten minutes.

So a two-minute breathing session at a transition point is reasonable as a way of marking a boundary and of slowing the breath for a while. It is not a demonstrated performance intervention, and the honest version of the advice says so.

The Shape and the Pattern

The breathing side starts with the name. The square is a diagram of the counts, not of anything the lungs do. Four equal phases — inhale, hold, exhale, hold — drawn as four equal sides, which is a tidy picture and a slightly misleading one, because it implies the geometry is the feature. The geometry is a consequence. What the pattern actually specifies is a single number applied four times.

That is worth saying plainly, because it means box breathing has exactly one parameter. Set the count to four seconds and the cycle is sixteen seconds long, which is 3.75 breaths a minute. Set it to six and the cycle is twenty-four seconds, or 2.5 a minute. Set it to three and you get twelve seconds, five a minute. The square is identical in every case; only the rate moves, which makes the count length the one decision box breathing actually asks of you.

Two panels. The left panel shows one cycle of box breathing as a square path with an arrowhead on each side marking the direction of travel. The top side is a four-second inhale and the bottom side a four-second exhale, both drawn at full weight; the two vertical sides are four-second holds drawn in a fainter weight. Text in the middle of the square reads sixteen seconds per cycle, and below it, eight of those sixteen seconds hold. The right panel is a horizontal bar chart of breaths per minute for the site’s four presets: 4-7-8 at 3.2, Box at 3.75, Relax at 6.0 and Energy at 10.0. A shaded band spanning six to ten breaths per minute is labelled six to ten a minute. The 4-7-8 and Box bars both stop short of the band; the Relax bar reaches its lower edge and the Energy bar its upper edge. The shape the name describes The rate each preset works out to Inhale 4 s Hold 4 s Exhale 4 s Hold 4 s 16 s per cycle 8 of those 16 s hold 6 to 10 a minute 4-7-8 Box Relax Energy 3.2 3.75 6.0 10.0 0 4 8 12 Breaths per minute (at one second per count)
Left: the standard four-count box, with the two airflow sides at full weight and the two retentions recessive, because nothing moves during them. Right: breaths per minute for each of this site’s four presets, computed from their counts at one second per count, against the six-to-ten-a-minute band that Russo et al. 2017 identify. The square is the same shape at any count length; the rate is not, which is why the rate is the parameter worth choosing.

Against the slow-breathing literature, which the article on 4-7-8 breathing on this site walks through in more detail, that rate is low. A 2017 review in Breathe concluded that what it calls autonomically optimised respiration appears to sit in the band of six to ten breaths a minute, and a 2022 review identifies a resonance near six. A four-second box, at 3.75 a minute, is below both.

Work the arithmetic the other way and the gap is sharper than the four-second version suggests. A square cycle is four counts long, so the rate is sixty divided by four times the count length: to land inside the six-to-ten band a square needs counts of one and a half to two and a half seconds. That is faster than box breathing is usually taught, and even there half of each cycle is still a retention. The equal-sided shape and the studied rate pull against each other.

The timer on this site fixes the Box preset at four seconds, with no count-length control, so its box is always sixteen seconds a cycle. Across the three session lengths of three, five and ten minutes that comes to roughly eleven, nineteen or thirty-eight cycles.

The Two Trials That Tested It

Box breathing is one of the few named patterns to have been tested against alternatives rather than against nothing, and it has lost both times.

A 2023 randomised controlled trial compared three five-minute daily breathing exercises against an equal period of mindfulness meditation over one month. The three were cyclic sighing, which emphasises prolonged exhalations; box breathing, defined there as equal durations of inhalation, retention and exhalation; and cyclic hyperventilation with retention. Breathwork as a whole improved mood and lowered respiratory rate more than meditation did, but the effect was carried by cyclic sighing — the exhale-focused pattern — and that is the one the authors single out as promising.

A 2025 experiment took a different approach, measuring eighty-four people across four conditions in a single session: square breathing (the same pattern under its other name), 4-7-8 breathing, and six breaths a minute at two different inhale-to-exhale ratios. Six a minute raised heart rate variability more than square or 4-7-8 breathing, with small to medium effects. Nothing moved blood pressure or mood meaningfully. The authors are blunt about the pattern in their opening line, describing square and 4-7-8 breathing as popularly promoted with little empirical support behind them.

Taken together: the pattern is well known, it has been examined, and in both examinations something simpler did better. That is a more useful thing to know than a list of its supposed benefits.

Equal Counts, Half of Them Still

The two retentions are what distinguish box breathing from an ordinary slow breath, and in the standard four-count version they occupy eight of the cycle’s sixteen seconds. Holding is not neutral. Breath retention lets carbon dioxide accumulate, which is the opposite of what happens when the breathing rate rises, and the 2025 experiment measured end-tidal carbon dioxide with a capnometer precisely because the authors expected the retention patterns to raise it. The published result they report on carbon dioxide is the unexpected one: breathing at six a minute produced mild over-breathing. What the retentions did to carbon dioxide is not stated in the summary of findings, so no claim about it belongs here.

What can be said is that no published study isolates the holds — lengthening them, shortening them, or removing them while keeping the rate fixed. The retentions are the pattern’s signature and its least examined component, which means confident explanations of what they accomplish are speculation wearing a mechanism’s clothes.

The equal counts do buy something real, and it is worth separating from the physiological claims. A pattern with one parameter is easy to execute without instrumentation: there is a single interval to keep in mind and no transition where the next phase differs in length from the last. Asymmetric patterns fail in a predictable direction when counted by hand, because the phase that gets shortened is the longest one, and in 4-7-8 the two longest phases are the hold and the exhale. A square has no longest phase to lose.

Set against that, the equal split carries a cost the shape conceals. Dividing a cycle into four equal parts and spending two of them holding means half the cycle moves no air, so a sixteen-second box delivers eight seconds of breathing. Slowing the breath by that route means spending much of the time not breathing at all, which is a different physiological situation from breathing slowly and continuously — and it is the situation nobody has isolated.

The equal phases also produce one small practical ambiguity. Both retentions carry the same label, so without a visual cue it is easy to lose track of whether the current hold follows the inhale or the exhale. On this site’s timer the circle settles it by size: expanded after the inhale, contracted after the exhale.

Choosing Box Over the Alternatives

Three questions settle it, and none of them needs a literature. Do you want a pattern you can run without looking at anything? The square has no longest phase to lose, so it survives being counted by hand better than 4-7-8 does. Do you want the rate the slow-breathing reviews work in? Then Box is the wrong preset and Relax is the right one, at four counts in and six out. Do you want the retentions themselves? Nobody has isolated them, so wanting them is a preference and worth holding as one. Box breathing remains a perfectly serviceable way to spend four minutes paying attention to the breath. It is just not the pattern the evidence points to.

References

Frequently Asked Questions

What is box breathing and why is it called that?

It is a breathing pattern with four phases of equal length: inhale, hold, exhale, hold. The name comes from drawing those four equal phases as the four sides of a square. The shape is a consequence of the equal counts rather than a feature in itself, which means the pattern has one parameter — the count length.

Does box breathing have evidence behind it?

It has been tested twice against alternatives and finished second both times. A 2023 randomised trial compared it with cyclic sighing and cyclic hyperventilation against meditation; the benefit was carried by cyclic sighing. A 2025 experiment compared it with 4-7-8 and with plain six-breaths-a-minute pacing; six a minute raised heart rate variability more, and no condition meaningfully changed blood pressure or mood.

What do the two breath holds do?

Nobody has established this. No published study varies or removes the holds while keeping the breathing rate fixed, so the pattern’s signature feature is also its least examined one. Holding does let carbon dioxide accumulate, which is why the 2025 experiment measured it, but the reported carbon dioxide finding concerns the six-a-minute condition rather than the retentions.

Is box breathing worth pairing with a pomodoro work block?

As a way of marking the boundary between work and break, yes — that costs two minutes and asks nothing of the evidence. As a performance intervention, the claim overreaches. A 2022 meta-analysis of micro-breaks found small reliable gains in vigour and reductions in fatigue but no significant effect on overall performance, with benefits confined to less cognitively demanding tasks.

WhiteNoise.top

WhiteNoise.top builds free browser-based noise generators, mixers and breathing timers. Our articles explain the acoustics behind those tools and cite published research where it exists.