Quick Answer
Slow breathing at about six breaths a minute has the most data behind it. Cyclic sighing is the best five-minute version. Cold water on the face gives the biggest shift you can measure in one session. All three are free. One placebo-controlled trial of slow breathing found no benefit over a convincing fake, so skip the devices.
Affiliate disclosure: This post contains affiliate links. See our affiliate disclosure for details.
We ranked these exercises by the strength of the evidence behind them, and three come out ahead: slow breathing, cyclic sighing and cold water on the face. The research on slow breathing calls it low-tech and low-cost. We own a cedar barrel sauna and a backyard ice bath with no chiller, so the cold part we know first-hand. We list what we own on our about page.
Last reviewed: October 2026
The Exercises, Ranked by What the Evidence Shows
| Exercise | What the evidence shows | Strongest study |
|---|---|---|
| Slow breathing (~6/min) | Raises vagal HRV during a session, right after one, and after a course | Review of 223 studies |
| Cyclic sighing | Beat daily mindfulness meditation on mood and breathing rate over a month | Randomised trial, 27 in the sighing arm |
| Cold water face immersion | Moderate to large rise in RMSSD while the face is cold | Review of 17 papers |
| Cold water face immersion in recovery | Faster recovery than sitting still | Trial, 13 men |
| Humming (bhramari) | RMSSD up in a 20-person pilot, uncontrolled design | Pilot study, 20 people |
| Coherent breathing, paced | No better than a well-designed fake over four weeks | Placebo trial, 400 people |
Ratings and ordering are ours. They weight the quality of the control group, not the size of the effect.
What a "Vagus Nerve Exercise" Is Actually Measuring
A review in Psychophysiology, which we come to further down, measures vagal activity with one number: RMSSD. It is built from the differences between adjacent heartbeat intervals. On this page, that number is what we mean by vagal HRV, short for heart rate variability.
Two things matter about it. First, your heart rate speeds up and slows down with your breathing, and that swing is biggest at about six breaths a minute. Second, in that Psychophysiology review the RMSSD rise was there during exposure and gone afterwards.
So there are two claims in play. One is that an exercise moves a body reading right now. The evidence for that is good. The other is whether doing it daily changes how you feel over four weeks, which is what the trials below tested. That evidence is thinner, and the control group starts to matter a lot. Our view: these are not the same claim, and the gadget marketing here depends on you not separating them. If you would rather watch your own numbers, we rank the straps in our guide to the best HRV monitors.
Slow Breathing: The One With 223 Studies Behind It
This is also the one nobody needs to sell you.
A 2022 review in Neuroscience & Biobehavioral Reviews screened 1,842 abstracts. It included 223 studies of voluntary slow breathing. It found rises in vagal HRV in all three windows it looked at: during the session, right after a single session, and after a course of sessions. The authors wrote that slow breathing "could be advised as a low-tech and low-cost technique to use in prevention and adjunct treatment purposes, with few adverse effects expected".
Lehrer and Gevirtz reviewed breathing with live heart rate feedback in Frontiers in Psychology. They put the rate this way: "most studies find that maximum effects usually are achieved when breathing at a rate of approximately 0.1 Hz (six breaths per minute)". The same review reports that taller people and men tend to have lower resonance frequencies. Six is the sensible default.
What we suggest: the rate the research names is about six breaths a minute. A 2021 study that paced it at that rate had people inhale continuously through the nose for four seconds and exhale continuously through pursed lips for six, on the grounds that a longer out-breath produces a bigger rise in vagal activity. That is the split we use. The same study compared sessions of five, ten, fifteen and twenty minutes and found no difference in RMSSD between them, so we stop at five. If you are straining for air, shorten the out-breath rather than toughing it out.
That is the whole technique.
Cyclic Sighing: The Best Five Minutes If You Only Do One Thing
Cyclic sighing is a breathing exercise that drags the out-breath out on purpose. The trial that tested it described two differences from the other breathing groups: a longer out-breath, and a double inhale that deepened the in-breath.
It got its own randomised trial in Cell Reports Medicine in 2023. The study ran remotely for a month. It compared three five-minute daily breathing exercises against five minutes of daily mindfulness meditation. Of 140 people who consented, 114 were invited into the study. The cyclic sighing arm had 27 people in the physiological analysis, against 22 in the meditation group. Cyclic sighing beat meditation on mood and on breathing rate, and both differences cleared the study's usual bar for statistical significance.
Two things make this one stand out. It was not measured against doing nothing: the comparison group did an equivalent period of daily mindfulness meditation. And five minutes is short enough that the habit survives a bad week.
Two things count against it. The sighing arm had 27 people in it. And Andrew Huberman is an author on the study, and he recorded the video instructions the participants followed. That does not make the result wrong. Our view: an independent repeat would help before anyone calls five minutes of sighing settled.
What we suggest: inhale through your nose. Take a second shorter sip of air on top to fill up. Exhale slowly through your mouth until you are empty. Repeat for five minutes. That second inhale is one of the two things the trial singles out.
Cold on the Face: The Biggest Same-Session Shift
That same Psychophysiology review is a review of the diving response. The triggers it examined were face immersion or cooling, SCUBA diving, and total body immersion into water. Of everything on this page, this is the effect we would bet on showing up.
That review pooled 17 papers covering 311 people. Those papers measured RMSSD around those triggers. During exposure it found a moderate to large rise while the face was cold. After exposure it found no clear effect. Full body immersion worked better than cooling the forehead alone.
A study in the European Journal of Applied Physiology looked at whether the effect carries over after hard exercise. It used 13 men, and each of them tried both recovery methods, so every man was his own comparison. Each did a hard interval bout twice. They then recovered for five minutes either with cold water face immersion or sitting still. The researchers report that vagal activity came back faster with the cold water face immersion. Thirteen men, and that is the honest size of the claim.
The cold is the one part of this we have hands on. Our ice bath has no chiller, so its temperature is whatever the weather and a bag of ice decide that morning. What stands out is that the face is where the sensation really is. Shoulders-in is unpleasant and manageable. The face is a different order of thing. You do not need a tub for this. Face immersion is one of the triggers that review examined. We go deeper into the bowl version in our piece on the ice bath for your face. The wider case for the tub is in our cold plunge benefits guide.
What we suggest: fill a bowl with cold water and add ice if you have it. Hold your breath and put your face in, temple to chin. Come up when you need a breath, breathe normally, then go again.
One caution worth the space. A review in The Journal of Physiology reports that cold water submersion can induce a high incidence of cardiac arrhythmias in healthy volunteers. Arrhythmias are electrical disturbances to the heart, in that review's own words. So we would ask a doctor first with any heart condition, and we would not do the tub version alone.
Humming: Real Data, Awkward Findings
A 2023 study in Cureus put Holter monitors on 23 people. It compared humming against physical activity, emotional stress and sleep. Humming produced the lowest stress index of the four.
A separate 2026 pilot in the International Journal of Yoga Therapy used 20 participants. Each did a single five-minute session of humming bee breath, known in yoga as bhramari pranayama. The researchers recorded both HRV and brain waves. The Psychophysiology review indexes cardiac vagal activity with RMSSD. In this pilot RMSSD rose from 29.95 ms to 45.83 ms, far past the usual bar for statistical significance.
The same paper reports something else. The brain-wave trace showed less delta and theta with more beta and gamma, which the authors call "a state of cortical alertness". Their own conclusion flags that "the uncontrolled design and small sample size limit generalizability".
Our read: the changes here are measurable, and the paper's own reading of the brain-wave trace is alertness rather than calm. We would not pay for a gargling or singing product either, and we would not do any of them instead of the breathing.
The Honest Ceiling: A Placebo Trial Found Nothing Extra
One trial ran coherent breathing against a matched placebo that participants rated just as believable. That is why we are not more enthusiastic about any of this.
It was published in Scientific Reports in 2023, and it calls itself the largest parallel randomised controlled trial on breathwork to date. It split 400 people between coherent breathing at about 5.5 breaths per minute and a matched attention-placebo at 12 breaths per minute. Both groups did about ten minutes a day for four weeks. Nobody knew which group they were in. Both groups breathed with the same in-to-out ratio. Both rated their own exercise equally believable.
On the main outcome, self-reported stress, the two groups did not differ over time. There was a significant main effect of time. Both groups improved, and kept improving at one-month follow-up, on stress, anxiety, depression and wellbeing. The authors' phrasing is exact: there was "no measurable effect of coherent breathing over and above a well-designed breathwork placebo at improving mental health and wellbeing".
That sits next to the review of breathing exercises, which found a small-to-medium effect on stress, pooled from 12 randomised trials and 785 adults. The authors urged "caution" and asked for "nuanced research approaches with low risk-of-bias study designs to avoid a miscalibration between hype and evidence".
These two results do not say breathing does nothing. People in both arms of that trial got better. Our read: sitting down may be doing most of the work, and the exact rhythm less. That is fine if all you want is to feel better. It is also a reason to think twice before paying for a device whose whole pitch is the rhythm.
Do You Need a Device?
No. Our answer is a clock and a bowl of water. If you are going to buy something anyway, here is what each option does.
| Product | What it adds | Price | Needed? | Rating |
|---|---|---|---|---|
| A clock and a bowl of water | Nothing, and nothing is missing | Nothing to buy | This is the baseline | - |
| moonbird | Tactile breathing guidance that expands and contracts in your hand, plus heart rate and HRV from an optical sensor | ~$180 | No | 4.0 |
| Polar H10 | A heart rate monitor chest strap, so you can see whether your own numbers move | ~$100 | No | 4.6 |
| Pulsetto FIT | Sold as a vagus nerve stimulation device | ~$270 | No | 3.8 |
| Apollo Neuro | A wearable plus a one-year SmartVibes AI membership | ~$450 | No | 3.4 |
Prices last checked on 11 November 2026 on moonbird.life, polar.com, pulsetto.tech and apolloneuro.com.
Confirm the current figure on the brand's own site before you buy.
moonbird is our pick of these, because it helps with the exercise rather than standing in for it. moonbird lists tactile breathing guidance that expands and contracts in your hand. It also lists an optical heart rhythm sensor measuring heart rate and HRV, and a two-hour battery life. Its own FAQ answers "do I need the app" with "No. Moonbird works as a standalone device with guided breathing exercises". That is a straighter answer than we expected. The device with the basic app is about $180 one-time, and there is a premium app tier you can subscribe to and cancel. moonbird's own product page also argues that "a generic pace doesn't cut it: the rhythm that activates your vagus nerve most effectively is unique to your body". The personalised pace it sells as the answer sits behind the Premium tier. Our read: that is very convenient.
The Polar H10 is the cheapest honest answer to "is this working for me". It is a chest strap at around $100. If you are going to spend money on this topic, spend it on measurement.
Pulsetto and Apollo Neuro are a different proposition. Our view: we would rather spend five minutes a day on the exercises above than buy either. We went through the published evidence for the neck devices in our Pulsetto review, and our verdict there was that the evidence is thinner than the marketing. Apollo's own store lists the wearable at about $450 with a one-year membership attached.
Our Verdict
Starting from nothing today, we would do five minutes of cyclic sighing every morning. We would add a cold face dunk on the days we use the ice bath anyway. And we would spend nothing. Our read: that pairing has the best evidence for the effort. The trial behind cyclic sighing beat daily mindfulness meditation on mood and on breathing rate.
If the habit will not stick without something in your hand, the moonbird at around $180 is the one we would buy. A brand whose FAQ tells you the app is optional has earned a little trust. If you would rather know whether any of it works, buy the Polar H10 for about $100 and watch your own RMSSD.
What we would not buy is stimulation. The placebo trial of slow breathing should govern spending here. Both arms improved and neither beat the other. Our read: the sitting down is doing more than the protocol. The Pulsetto FIT lists at about $270 on Pulsetto's own store, and we would keep the money. We also have a guide to using a sauna for stress relief, and the rest of our recovery coverage sits alongside it.
FAQ
What are the best vagus nerve exercises?
Slow breathing at about six breaths a minute, cyclic sighing, and cold water on the face. A 2022 review of 223 studies found slow breathing raised vagal HRV during a session, right after one, and after a course of them. Cyclic sighing beat daily mindfulness meditation on mood in a randomised trial. Cold on the face produced a moderate to large rise in RMSSD in a review of 17 papers. The slow-breathing research calls the technique low-tech and low-cost.
How do you stimulate the vagus nerve naturally?
Slow your breathing down and make the out-breath longer than the in-breath. Or cool your face. Both raised vagal HRV in the trials above. The slow-breathing review pooled 223 studies; the face-cooling review pooled 17 papers. Lehrer and Gevirtz report that most studies find maximum effects at about six breaths a minute.
How long does it take for vagus nerve exercises to work?
In the Psychophysiology review the effect on RMSSD was there while the face was cold and gone after it. For how you feel, the cyclic sighing study ran daily five-minute exercises over a month, and sighing beat meditation on mood. The coherent breathing trial ran about ten minutes a day for four weeks.
Does humming stimulate the vagus nerve?
Two small pilot studies measured HRV during humming. In the 20-person pilot RMSSD rose from 29.95 ms to 45.83 ms. The Cureus study reported humming produced the lowest stress index of the four activities it compared. The pilot's authors say its "uncontrolled design and small sample size limit generalizability". It also recorded more beta and gamma brain activity, which its own authors called a state of cortical alertness. Our read: the changes are measurable, but calm is not what the authors read off that trace.
Do vagus nerve stimulation devices work?
Not on the evidence we found. The trial we lean on hardest calls itself the largest parallel randomised trial of breathwork to date. It split 400 people between coherent breathing and a matched fake rhythm for four weeks, and found no difference between them on stress, anxiety, depression or wellbeing, even though both groups improved. Our read: if a convincing fake can match a real breathing protocol, a neck device needs better evidence than its marketing before we would spend on it. We went through what has been published for the neck devices in our Pulsetto review.
What is the 4-6 breathing technique for the vagus nerve?
Inhale through your nose for four seconds, exhale through pursed lips for six, continuously at both ends. That is about six breaths a minute, the rate Lehrer and Gevirtz's review of breathing with live heart rate feedback names as the one producing maximum effects. A 2021 study paced slow breathing at that rate with exactly that split, because a longer out-breath was found to produce a larger rise in vagal activity. It compared five, ten, fifteen and twenty minute sessions and found no difference in RMSSD between them. If you find yourself straining for air, shorten the out-breath rather than pushing through.
Can you overdo vagus nerve exercises?
Slow breathing has few reported bad effects, and the 223-study review said as much. Our view: cold water on the face is the part that needs care. A review in The Journal of Physiology reports that cold water submersion can induce a high incidence of cardiac arrhythmias in healthy volunteers. We would ask a doctor first with any heart condition, and we would not do the tub version alone.
Our Top Pick
moonbird
From ~$180
Before you buy
Where the premium is materials, not marketing
We take the spec sheets apart across saunas, cold plunge, red light and recovery, and show where the flagship price buys real hardware, and where it just buys the logo.
Disclosure: This article contains affiliate links. If you purchase through these links, Bankroll Zen may earn a commission at no extra cost to you. We only recommend products we genuinely believe in. Learn more in our affiliate disclosure.