CNS Oxygen Toxicity

The poison you can't live without

Every diver gets taught that running out of air is what kills you. Almost nobody gets taught that too much oxygen can do it just as easily. Breathe oxygen at a high enough pressure and it can set off a seizure underwater, with a regulator still in your mouth and little or no warning. The gas itself hasn't changed at all. What's changed is the pressure you're breathing it at, and past a certain point oxygen stops being fuel and starts acting like a poison.

For a diver, the scary version of that is CNS oxygen toxicity: a sudden seizure that can hit with no warning, while you have a regulator in your mouth and water all around you. The seizure itself almost never kills anyone. Drowning during the seizure does.

The good news is that this is one of the most preventable problems in diving. You don't avoid it by being brave or experienced. You avoid it with arithmetic: know your oxygen pressure, and keep it inside limits we already know. The rest of this article is where those limits come from and how to stay on the right side of them.

One quick calibration before we go further. If you dive recreational nitrox, this is mostly a planning discipline you'll rarely bump up against (EAN32 at 30 m is only a PO₂ of 1.28). The sharp end of this article, where seizures actually happen, is rebreather and decompression diving. The arithmetic is the same either way, so it's worth owning now.

Why oxygen turns toxic

Your body runs on oxygen, but it also has to defend against it. Burning oxygen for energy throws off a small amount of chemical shrapnel: unstable molecules called reactive oxygen species (ROS). Think of them as sparks coming off a grinding wheel. At normal pressure your body's antioxidant systems sweep those sparks up as fast as they appear.

Now raise the pressure. The deeper you go, the more oxygen dissolves into your blood and tissues, far more than your body was ever tuned for. The sparks start flying faster than the cleanup crew can handle, and the leftovers begin damaging cells. That's oxidative stress, and it's the root of every form of oxygen toxicity.

One leading explanation of the seizure itself is that, in the brain, this oxidative stress tips the chemistry toward over-excitement. It eats into GABA, the signal your neurons use to calm each other down, while nudging up the excitatory side. GABA is the brakes. Ease the brakes and press the accelerator at the same time and the neurons start firing together, faster and faster, until the whole system runs away. That runaway is the seizure. The exact mechanism is still argued over, but the practical lesson holds either way.

The one number that matters: PO₂

Depth alone doesn't poison you. Your oxygen mix alone doesn't either. What matters is the two multiplied together, the partial pressure of oxygen, or PO₂. It's just the share of the pressure around you that is oxygen:

PO₂ = oxygen fraction × absolute pressure

Absolute pressure is easy in seawater: every 10 metres adds one bar, and you start at 1 bar at the surface. So at 30 m you're at 4 bar. Breathe air (21% oxygen) there and your PO₂ is 0.21 × 4 = 0.84 bar. Breathe EAN32 (32% oxygen) at the same depth and it's 0.32 × 4 = 1.28 bar. Same depth, very different oxygen dose, which is exactly why a richer mix has a shallower limit.

Now the part most courses skip: 1.4 and 1.6 are not a physical cliff. Nobody convulses at 1.59 and walks away at 1.61. In dry chamber tests, resting subjects can sometimes tolerate 2 bar or more before seizing. The in-water limits sit far below that for two reasons: immersion, exercise, cold, and carbon dioxide all drag the real threshold down, and the cost of guessing wrong is a drowning. So treat 1.4 and 1.6 as deliberately cautious safety margins, not magic numbers. People have seized below 1.6, and plenty have dived above it without incident. Risk here is a dial, not a switch.

Turning the limit into a depth (MOD)

Flip the PO₂ formula around and it tells you the deepest you can take a given mix, its Maximum Operating Depth (MOD):

MOD (m) = ( PO₂ limit ÷ oxygen fraction − 1 ) × 10

The −1 is there because your pressure starts at 1 bar at the surface, the same fact you used in the PO₂ formula above.

Two things to notice. First, those air numbers are academic: long before air reaches its oxygen limit you'd be hammered by nitrogen narcosis, which is why nobody actually dives air anywhere near 57 m. Second, EAN50 and 100% O₂ are deco gases, rich mixes breathed only on shallow decompression stops, never on the bottom, and there's barely any room between their working MOD and their ceiling: for EAN50 it's just 4 metres. Drift down past your stop and you can blow through 1.6 without noticing, which is why deco-gas depths get planned to the metre.

Dose adds up over time: the "CNS clock"

A single PO₂ reading only tells you the intensity of the exposure, not how long you've been soaking in it. Two hours at a high PO₂ is harder on you than two minutes. So divers also track a running total, usually called the CNS clock or CNS %.

The idea is simple. For every PO₂, there's a published time limit (from the NOAA tables). Spend half that time and you've used 50% of your "budget" for that level; your computer adds up the fractions across the whole dive and shows one number. 100% means you've reached the recommended single-exposure limit, a planning red line, not the moment you seize.

*The 1.3 row is the old figure. It was revised sharply upward in 2025, see the note below.

Between dives the clock winds back down (computers usually halve it about every 90 minutes), so a decent surface interval restores most of your budget.

A whole dive, end to end

Put both tables together on one dive. Say you plan EAN32 to 32 m for 30 minutes, then switch to EAN50 at 18 m for 10 minutes of deco.

• Gas check. EAN32's working MOD is about 34 m, so 32 m is fine (PO₂ there is 0.32 × 4.2 = 1.34 bar). EAN50's working MOD is about 18 m, so you switch right at its limit (PO₂ 0.50 × 2.8 = 1.40 bar). Both legal, nothing over 1.4.
• CNS clock. Thirty minutes at about 1.34 bar spends roughly 30 ÷ 168 ≈ 18%. Ten minutes at 1.40 bar spends 10 ÷ 150 ≈ 7%. Total ≈ 25%, comfortably under the 80% you'd plan to.

So this dive is limited by your gas supply and your decompression, not by oxygen, which is the usual story on a single nitrox dive. Stack four or five of these on a liveaboard day, though, and the clock starts to matter, which is exactly what the tool above lets you feel.

The warning signs that don't warn you

You'll be taught a mnemonic, VENTID-C, for the symptoms that can come before a seizure:

• V is for Vision (tunnel vision, blurring, flashes)
• E is for Ears (ringing)
• N is for Nausea
• T is for Twitching, classically of the lips and face
• I is for Irritability or sudden anxiety
• D is for Dizziness
• C is for Convulsion

Learn it, but do not trust it. This is the single most important sentence in the article.

Carbon dioxide: the hidden accelerator

If you remember one risk factor, make it this one. Carbon dioxide is the master switch that makes oxygen far more dangerous. When CO₂ builds up in your blood, it widens the blood vessels in your brain, pushing more blood, and therefore more oxygen, into exactly the tissue you're trying to protect. The effective oxygen dose to your neurons goes up even though your PO₂ on the computer hasn't changed.

CO₂ builds up when you work hard, when you breathe a dense gas that's tiring to move, when you "skip-breathe" (deliberately pause between breaths) to stretch your gas, and, on a rebreather, when the scrubber that removes CO₂ starts to fail. Scrubber breakthrough has triggered seizures at a PO₂ as low as 1.2 to 1.3 bar [3]. On a rebreather, managing CO₂ is managing oxygen toxicity. There's a whole article on that side of it: gas density and CO₂ buildup.

Everything else that raises your risk works through a similar door. Hard exertion, cold, and simply being immersed in water all lower the seizure threshold. Susceptibility varies between people too, and the same diver isn't equally tolerant every day.

If your buddy convulses

You cannot stop the seizure, so your whole job is to prevent a drowning and a burst lung:

1. Hold them at depth and stop the ascent. A clenched, breath-holding body wants to float up, and a held breath on the way up can rupture a lung.
2. Leave the regulator alone during the convulsion. If it's in the mouth, hold it there. If it has already come out, do not try to force it back against a clenched jaw; you can't, and you may cause injury.
3. Don't try to pin the limbs. You won't stop the convulsion and you'll just hurt yourself.
4. Wait for the shaking to stop. Do not ascend during the active convulsion.
5. Once it passes and they're breathing, replace the regulator if it came out, make a slow, controlled ascent together, then get them onto oxygen at the surface and to emergency medical care.

The honest summary

CNS oxygen toxicity sounds terrifying, and the consequence genuinely is. But the cause is boringly mechanical: too much oxygen pressure, often pushed over the edge by carbon dioxide. You control the first with a mix and a depth limit you can calculate before you ever get wet, and the second with fitness, good gear, and calm breathing. Do both and the risk drops to a tiny fraction of a percent [4]. Skip the arithmetic and trust your symptoms to save you, and you're betting your life on a warning system that frequently doesn't fire.

Keep reading

• Oxygen Toxicity & ROS: what the damage actually is
• Pulmonary Oxygen Toxicity: the slow lung version
• Oxygen on a Multi-Day Trip: what the 1.3 rule means over a week
• Gas Density & CO₂ Buildup: the hidden accelerator, in depth

References

1. Hoyt J, Murphy FG, Mitchell SJ, et al. Revised guideline for central nervous system oxygen toxicity exposure limits when using an inspired PO₂ of 1.3 atmospheres (≈1.3 bar). Diving and Hyperbaric Medicine. 2025;55(3). PMID 40986922.
2. Donald KW. Oxygen and the Diver. SPA Ltd, 1992. Summarising the Royal Navy oxygen-tolerance experiments, 1942–1945.
3. Pollock NW. Hazards in Rebreather Diving. 2024.
4. Shearwater Research. Oxygen Seizures at PO₂ ≤ 1.6 bar: how rare? A review putting reported CNS-toxicity incidence in clean conditions well under 1 in 1,000 dives.
5. Arieli R, et al. CNS oxygen toxicity in closed-circuit diving: symptoms reported from 2,527 dives. Aviat Space Environ Med. 2006. PMID 16708533.
6. NOAA. NOAA Diving Manual (source of the CNS oxygen-exposure limits and the 1.4/1.6 conventions).