- Four in through the nose, seven held, eight out through the mouth. One cycle takes 19 seconds, which is 3.16 breaths a minute against a resting rate of 12 to 20.
- The exhale is twice the length of the inhale, and that ratio is the part doing the work — measurably slowing heart rate and dropping blood pressure while you do it.
- In the one randomized trial that pitted it against ordinary deep breathing, 4-7-8 reduced anxiety more.
- The sleep claim it is famous for is the weakest thing on the page. One uncontrolled study in COPD patients is not the same as evidence that it puts you to sleep.
- Andrew Weil’s own instruction is four cycles and no more for the first month. Ignoring that is the most common way people give themselves a headache.
What 4-7-8 Breathing Is
Inhale quietly through the nose for four. Hold for seven. Exhale through the mouth for eight, with an audible breath out. That is one cycle, it takes nineteen seconds, and it works out at 3.16 breaths a minute — roughly a quarter of a normal resting rate.
Andrew Weil brought the pattern into general circulation in the early 2000s, drawing on pranayama, and he calls it a natural tranquilizer for the nervous system. Two details from his instruction get dropped almost everywhere it is repeated, and both matter. The tongue rests against the ridge of tissue behind the upper front teeth throughout. And you do four cycles, no more, for the first month.
What separates 4-7-8 from most slow-breathing patterns is the ratio rather than the rate. The exhale is twice the inhale, and there is a long hold in between. That asymmetry is the whole design: an exhale-weighted breath is the one shape the research consistently associates with a shift toward parasympathetic activity, and this pattern pushes it further than almost anything else in common use.
How to Do 4-7-8 Breathing
- Put the tip of your tongue behind your upper front teeth. On the ridge of tissue, not the teeth themselves. It stays there for the whole exercise, including through the exhale, which is why the breath out is a whoosh rather than a sigh.
- Empty your lungs completely through the mouth. Before the first count. Starting a nineteen-second cycle on top of air you never let go of is why people run out on the exhale.
- Inhale quietly through the nose for four. Quietly is part of the instruction. If you can hear the inhale, it is too fast or too big.
- Hold for seven. Throat relaxed, airway open. This is the longest phase and the one people shorten without noticing.
- Exhale through the mouth for eight, audibly. Steady the whole way out, not a dump followed by a wait. The count matters less than the fact that it is twice as long as the inhale.
- That is one cycle. Do four. Four cycles is 76 seconds. Weil’s instruction is to stay at four for the first month before going to eight, and there is a reason for it — see the safety note below.
The counting is where this falls apart unaided. Nineteen seconds is long enough that most people speed up without realizing, and the seven-count hold is the first thing to compress. The timer in the sidebar paces all three phases, and the full timer adds sound cues so you can do it with your eyes closed.
What the Evidence Actually Shows
4-7-8 has been tested more directly than most named breathing patterns, which is unusual and useful. It has also been marketed far ahead of what those tests found. Each claim below is rated against research that tested that claim.
| Claim | Verdict | What the Research Actually Says |
|---|---|---|
| It slows heart rate and lowers blood pressure while you do it | Supported | In 43 healthy young adults, a single session of 4-7-8 breathing lowered heart rate by 7.2% and systolic blood pressure by 3.8% in well-rested participants, with high-frequency HRV power rising. The effects were smaller but still present in sleep-deprived participants.1 |
| It reduces anxiety, and more than plain deep breathing | Supported | A randomized controlled trial in 90 post-surgical patients compared 4-7-8, ordinary deep breathing, and no intervention. State anxiety after the intervention was significantly lower in the 4-7-8 group than in either the deep breathing group or the control group.2 A meta-analysis of breathwork trials generally finds small-to-medium reductions in anxiety.5 |
| It improves sleep quality | Qualified | One study measured it: 45 COPD patients, sleep quality scores falling from 13.3 to 4.9 on the Pittsburgh index.4 That is the number the headlines quote. It was an uncontrolled before-and-after study in a clinical population, with no comparison group, so the improvement cannot be attributed to the breathing. We could find no controlled trial testing 4-7-8 for sleep in people without a respiratory condition. |
| It raises heart rate variability more than other slow patterns | Not Shown | A study of 84 students compared square breathing, 4-7-8, and two six-breaths-a-minute patterns head to head. Breathing at six a minute raised HRV more than either 4-7-8 or the square, with small to medium effects. It also found no meaningful change in mood or blood pressure in any condition.3 |
| It puts you to sleep in about a minute | Not Shown | We could find no study measuring how long 4-7-8 takes to produce sleep. The sixty-second framing comes from the length of four cycles, which is 76 seconds — a description of the exercise, not a finding about what it does. |
The Sleep Claim, Examined
Almost every page about 4-7-8 leads with sleep, so it is worth being precise about what exists. One study has measured 4-7-8 against sleep quality. Forty-five COPD patients at a single hospital practiced it, and their scores on the Pittsburgh Sleep Quality Index fell from 13.3 to 4.9.4 That is a large improvement, it reached significance, and it is the source of the numbers currently circulating in the press.
There was no control group. Everyone in the study got the breathing exercise, so there is nothing to compare the improvement against — not a placebo, not a waiting list, not another technique. People enrolled in a sleep study while sleeping badly tend to improve regardless of what you give them, partly through regression to the mean and partly through the attention itself. A drop of eight points on the Pittsburgh index in three weeks is also very large for a breathing exercise, which is a reason for more caution rather than less.
None of this means it does not work. It means one uncontrolled study in people with a respiratory disease is not a basis for telling a healthy person with insomnia that this will put them to sleep.
What can be said with more confidence is the mechanism-shaped version of the claim: a single session measurably lowers heart rate and blood pressure,1 and reduces state anxiety more than ordinary deep breathing does.2 If what keeps you awake is a racing mind and an activated body, an intervention that reliably lowers arousal is a reasonable thing to reach for at bedtime. That is a good argument. It is not the same as a demonstrated sleep effect, and the two are worth keeping apart.
Why It Works
Slow breathing with a long exhale increases the vagus nerve’s influence on the heart. Systematic review of the slow-breathing literature finds consistent rises in heart rate variability and respiratory sinus arrhythmia, along with EEG changes and self-reported calm — while noting the evidence is limited, the techniques studied are heterogeneous, and none of the included studies were randomized controlled trials.6
Two things are specific to this pattern. The first is the 2:1 exhale-to-inhale ratio, which is more exhale-weighted than almost any pattern in common use, and exhale weighting is the feature most consistently associated with the parasympathetic shift. The second is the seven-count hold, which raises carbon dioxide slightly. That is the phase most likely to make you lightheaded, and it is also the reason Weil caps the first month at four cycles.
At 3.16 breaths a minute, 4-7-8 runs well below the six-a-minute rate the slow-breathing literature converges on — the resonance frequency where the baroreflex produces the largest heart-rate oscillations. When the three were compared directly, six a minute raised HRV more than 4-7-8 did.3 If the physiological marker is what you are after, a plain even count at six a minute is the better tool. If what you want is something that reliably calms you down before bed, the ratio matters more than the rate.
When Something Else Is the Better Choice
- You want the highest HRV response. Six breaths a minute beat 4-7-8 on every HRV measure in the one head-to-head comparison.3 The coherent breathing pacer runs exactly that.
- The seven-count hold makes you anxious or lightheaded. Drop the hold and keep the ratio: four in, eight out, no hold. You lose the CO₂ rise and keep the part that does the work. The extended exhale timer paces it.
- You need it to work in seconds, not a minute. A physiological sigh — double inhale, long exhale — is the fastest route to a drop in arousal and takes one breath.
- You have a respiratory condition. Breath holds and forced exhalation are not neutral in asthma or COPD. Talk to whoever manages it before adding this.
Variations Worth Knowing
| Pattern | Rate | Use it when |
|---|---|---|
| 4-7-8 | 3.16 breaths a minute | The standard. Four cycles, twice a day, for the first month |
| 4-4-6 | 4.3 breaths a minute | The full pattern is too long. Keeps the exhale weighting, shortens the hold |
| 4-0-8 | 5 breaths a minute | The hold is the problem. Same ratio, no CO₂ rise, no lightheadedness |
| 2-3.5-4 | 6.3 breaths a minute | Halved counts. Where to start if the full version leaves you gasping on the exhale |
All four keep the exhale longer than the inhale, which is the one thing not worth changing. What you can safely adjust is the absolute length and the hold.
Safety and the Four-Cycle Rule
Weil’s instruction is four cycles maximum for the first month, then eight. That cap is the single most-ignored part of the technique and the reason most people who report headaches or dizziness got them. The seven-count hold followed by a forced eight-count exhale moves more air and more carbon dioxide than people expect, and doing twenty cycles on your first attempt is not enthusiasm, it is overbreathing.
Never practice breath holds in or near water, and not while driving. Talk to a doctor first if you are pregnant, or if you have a respiratory condition such as asthma or COPD, a cardiovascular condition, uncontrolled blood pressure, epilepsy or a seizure history, glaucoma, or panic disorder. Lightheadedness, tingling, or rising panic mean stop and shorten the counts — not push through.
Building the Habit
Twice a day is the recommended schedule, and the two uses are different. The acute one is for a specific moment — the ten minutes before sleep, after a difficult conversation, before something you are dreading. That is the use the anxiety evidence supports.2 The other is maintenance: doing it daily whether or not you feel you need it, on the theory that the baseline moves. That version has thinner support for this specific pattern, though it holds up better for slow breathing generally.5
Practically: attach it to something you already do at a fixed point rather than a time of day. Lights off and then four cycles is a better rule than nine o’clock. Four cycles is 76 seconds, which is short enough that the excuse for skipping it never really holds.
Common Questions
Does 4-7-8 actually help you fall asleep?
The honest answer is that we could find no controlled trial of it. One uncontrolled study in COPD patients found large improvements in sleep quality, but with no comparison group that number cannot be attributed to the breathing. What is better established is that a single session lowers heart rate, lowers blood pressure and reduces anxiety — which is a reasonable thing to want at bedtime, and a different claim from a demonstrated sleep effect.
Why only four cycles?
Because the seven-count hold and the forced eight-count exhale shift more carbon dioxide than people expect, and the commonest way to end a first session with a headache is to do twenty cycles because it felt easy at the time. Weil’s own instruction is four for the first month, then eight. We could find no study testing whether more is better.
Is 4-7-8 better than box breathing?
For anxiety, probably, on the reasoning that exhale-weighted patterns outperform symmetrical ones — though no trial has compared those two on anxiety directly. For heart rate variability, neither wins: when they were tested head to head alongside six-breaths-a-minute patterns, plain six a minute beat both. For getting people to keep doing it, box breathing has the easier count.
What is the tongue position for?
It is part of Weil’s original instruction, borrowed from pranayama, and it is what makes the exhale a whoosh rather than a sigh. There is no evidence it changes the physiological effect. It does give you something to check when your attention wanders, which may be the practical value of it.
Can I use it during a panic attack?
Be careful. Breath holds can make panic worse for some people, because air hunger is one of the sensations panic feeds on. If you want a breathing tool for acute panic, an exhale-focused pattern with no hold is safer to start with. Recurrent panic attacks are worth talking to a clinician about rather than managing alone.
Do the counts have to be seconds?
Weil says the ratio is what matters, and that you can go faster if the full length is uncomfortable — halving everything keeps 4-7-8 intact at 2-3.5-4. What you should not do is shorten the exhale to keep up, since the long exhale is the part doing the work.
The Short Answer
4-7-8 is one of the better-evidenced breathing patterns for the thing it is least often sold on. A single session demonstrably lowers heart rate and blood pressure, and it beat ordinary deep breathing on anxiety in a randomized trial — which is more than most named techniques can show. The sleep claim that made it famous rests on one uncontrolled study in a clinical population, and the sixty-second framing is a description of the exercise rather than a finding. Use it to come down from an activated state, at bedtime or anywhere else, and keep to four cycles until it is easy.
Four in, seven held, eight out, paced for you. Free, no account, nothing saved.
- Vierra, J., Boonla, O., and Prasertsri, P. (2022). Effects of sleep deprivation and 4-7-8 breathing control on heart rate variability, blood pressure, blood glucose, and endothelial function in healthy young adults. Physiological Reports, 10(13), e15389. doi:10.14814/phy2.15389
- Aktaş, G. K., and İlgin, V. E. (2023). The effect of deep breathing exercise and 4-7-8 breathing techniques applied to patients after bariatric surgery on anxiety and quality of life. Obesity Surgery, 33(3), 920–929. doi:10.1007/s11695-022-06405-1
- Marchant, J. (2025). Comparing the effects of square, 4-7-8, and 6 breaths-per-minute breathing conditions on heart rate variability, CO₂ levels, and mood. MS thesis, Brigham Young University. scholarsarchive.byu.edu/etd/10696 — a master’s thesis rather than a peer-reviewed paper, and the only head-to-head comparison of these patterns we are aware of. Weighted accordingly.
- Satria, O., Nasution, N., Apriani, F., Andriani, R., Damayanti, S., and Tarigan, A. P. (2025). Effectiveness of the 4-7-8 breathing technique in enhancing sleep quality for COPD patients. Jurnal Respirologi Indonesia, 45(4), 267–271. doi:10.36497/jri.v45i4.948 — quasi-experimental, pretest and posttest, no control group.
- Fincham, G. W., Strauss, C., Montero-Marin, J., and Cavanagh, K. (2023). Effect of breathwork on stress and mental health: a meta-analysis of randomised-controlled trials. Scientific Reports, 13, 432. doi:10.1038/s41598-022-27247-y
- Zaccaro, A., Piarulli, A., Laurino, M., et al. (2018). How breath-control can change your life: a systematic review on psycho-physiological correlates of slow breathing. Frontiers in Human Neuroscience, 12, 353. doi:10.3389/fnhum.2018.00353
Jay Newmark is a health writer with a background in kinesiology and seven years covering sleep, stress, and performance science for digital health and fitness publications. He writes about breathwork from a science-first perspective.