Ujjayi Breathing: What the 'Ocean Breath' Research Actually Says

At Breath Hold we talk a lot about control — how we breathe, how we manage the mind, how we handle stress. Ujjayi, the yogic "Ocean Breath", comes up constantly in that conversation.

I used to make two specific scientific claims about it in this article. I went back and read the papers. One of them I had backwards. So this is a corrected version: what Ujjayi is, what the studies genuinely found, and where it fits in breath-hold training once you strip out the folklore.

What Ujjayi is

Ujjayi (oo-JAI) is slow nasal breathing with a light constriction of the glottis, the opening at the back of the throat. That constriction adds resistance to airflow and produces the soft ocean-wave sound the technique is named for.

• Slow and steady
• Nose only, in and out
• Audible, but soft
• Deep without being forced
• Often with a small pause between breaths

The correction: the "52% oxygen boost" is not a benefit

I previously wrote that a 1991 study found experienced practitioners doing Ujjayi with a short breath hold got "a 52% boost in oxygen uptake", and that this was great news before a breath hold.

The number is real. My interpretation was wrong, and wrong in a way that matters.

The study (Telles & Desiraju, Indian Journal of Medical Research, 1991) measured oxygen consumption — how much oxygen the body was burning — not how much oxygen was being taken on board. Ten experienced practitioners were split by their kumbhaka (breath retention) style. The short-kumbhaka group, retaining about 22% of the breath cycle, showed a statistically significant 52% increase in oxygen consumption and metabolic rate. The long-kumbhaka group, retaining about 50% of the cycle, showed a 19% decrease.

For a breath-hold practitioner, a 52% rise in metabolic rate before a hold is the last thing you want. Higher oxygen consumption means you burn through your stores faster. Parkes' review of breath-hold physiology lists increased metabolic rate among the things that shorten breath-hold duration.

So if there is a lesson in that paper for us, it is the opposite of what I wrote: the long-retention variant, the one that lowered metabolic rate by 19%, is the interesting one. And even that is ten men in a 1991 lab, not a training prescription.

The correction: Ujjayi did not beat other slow breathing on oxygen saturation

I also wrote that a 2012 study found Ujjayi produced "the biggest jump in blood oxygen levels" compared to other slow-breathing methods.

The study I was thinking of is Mason et al., published in Evidence-Based Complementary and Alternative Medicine in 2013. It compared breathing patterns at 6 breaths per minute, with and without Ujjayi, in 17 yoga-naive participants. What it actually found:

• The largest improvements in baroreflex sensitivity and the largest drop in blood pressure came from plain slow breathing with equal inhale and exhale — not from Ujjayi.
• Ujjayi produced a similar increase in oxygen saturation, but slightly less improvement in baroreflex sensitivity and no change in blood pressure.
• Ujjayi on both inhale and exhale was the one pattern that did not significantly increase baroreflex sensitivity at all.
• The authors attributed Ujjayi's effects to raised intrathoracic pressure requiring greater effort than ordinary slow breathing, and flagged that the extra effort may stimulate sympathetic activity.

Baseline saturation in these participants was already about 98.3%, so there was very little headroom for anything to improve. Chasing SpO₂ gains in a healthy person at sea level is chasing a rounding error.

The correction: Ujjayi does not reliably raise your CO₂

I claimed Ujjayi slows breathing, which raises CO₂ retention, which builds tolerance. That mechanism sounds right and is not what the data show.

In the Mason study, the rise in oxygen saturation during slow breathing came with a moderate increase in ventilation and a drop in end-tidal CO₂. Slow breathing does not automatically mean less ventilation — if tidal volume grows to compensate, minute ventilation holds steady or rises and CO₂ goes down, not up. The same pattern shows up in slow-breathing studies at altitude: SpO₂ up, tissue CO₂ down, with tidal volume increasing and minute ventilation unchanged.

CO₂ only accumulates when you genuinely under-ventilate relative to metabolic demand. That is what a breath hold does. It is not something you can assume from the shape of a breathing pattern.

Safety consequence, and this is the important part: any breathing practice that lowers your CO₂ before a hold behaves like mild hyperventilation. It delays the urge to breathe without adding meaningful oxygen. That is the exact mechanism behind hypoxic blackout. Do not use Ujjayi, or any deep or effortful breathing pattern, as your final preparation before a breath hold. Read the breath-hold safety guide before you train.

What is actually well supported

Strip out the overreach and there is still something worth practising here.

Slow breathing works, and Ujjayi is a good way to get there. Breathing at around 6 breaths per minute raises cardiac-vagal baroreflex sensitivity and lowers blood pressure. A study using pharmacological blockade found the heart-rate variability changes during slow-paced yogic breathing are almost entirely vagally mediated — parasympathetic, not sympathetic. That is a real, mechanistically clean finding.

Ujjayi's practical advantage is control, not gas exchange. The Mason authors noted that the glottal resistance gives you finer control over airflow and therefore over breath rate. For someone who normally breathes at 12–18 breaths per minute, the sound and resistance make a 5–6 breath-per-minute pace much easier to hold. The audible feedback is a metronome you carry with you.

That reframing is smaller than "52% oxygen boost", and it is the one I can actually defend.

How to practise it

1. Inhale slowly through the nose while gently narrowing the throat. You should hear a soft "sss" — the Darth Vader sound.
2. Exhale through the nose, keeping the constriction, making a soft "hhh".
3. Keep the breath deep but relaxed, same pace and pressure throughout.
4. Aim for roughly equal inhale and exhale at about 6 breaths per minute. The Mason study found equal-ratio slow breathing gave the best autonomic response.

Two to three minutes is plenty to start. Good places for it:

• As a general daily practice for calm and vagal tone
• During recovery after a table, once your breathing has already returned to normal
• Warm-ups that are well separated from any hold

And one place it does not belong: the last minute before a breath hold. Keep that breathing calm, quiet, and unremarkable.

Final thought

Ujjayi is a good practice with a mediocre evidence base that has been badly oversold — including by me, in the earlier version of this article. It is a reliable route into slow breathing, and slow breathing has decent evidence behind it. That is enough. It does not need the invented numbers.

Got questions, or want to tell me how Ujjayi is going for you? Tag us on Instagram @breathhold.co.

Sources

• Oxygen consumption during pranayamic type of very slow-rate breathing (Telles & Desiraju, 1991)
• Cardiovascular and Respiratory Effect of Yogic Slow Breathing in the Yoga Beginner: What Is the Best Approach? (Mason et al., 2013)
• Vagal Mediation of Low-Frequency Heart Rate Variability During Slow Yogic Breathing, Psychosomatic Medicine
• Effects of Slow Deep Breathing at High Altitude on Oxygen Saturation, Pulmonary and Systemic Hemodynamics, PLOS One
• Breath-holding and its breakpoint (Parkes, 2006), Experimental Physiology
• Divers Alert Network: Hypoxia in Breath-Hold Diving
• Divers Alert Network: Freediving Safety