New recommendation for oxygen limits

Breathing oxygen for longer?

NOAA workshop considers longer exposure times at 1.3bar pO₂ unproblematic

The vast majority of divers associate oxygen primarily with one thing: a pO₂ limit. The exposure times appear to be of secondary importance, as even on a diving safari using nitrox, the current limits of oxygen exposure are hardly ever approached.
The situation is different on long rebreather dives, and even more so when really long deco times are added. In this area, dives are regularly undertaken in which 1.3 bar pO₂ is not exceeded as a maximum, but the maximum exposure times specified by the NOAA are certainly exceeded.

If qualified divers regularly disregard rules, it could be because the rules are too restrictive for that specific application, or the benefits of ignoring them are significant enough to make the “rule violation” widely accepted. However, because this constant disregard for rules has unpleasant side effects, it was worthwhile to review these rules again. To this end, at the beginning of the year, well-known figures from technical diving and hyperbaric medicine, under the leadership of Simon Mitchell, came together for a NOAA workshop and developed a new consensus. Simon Mitchell presented this new consensus in May 2025 at the annual meeting of the “South Pacific Underwater Medicine Society (SPUMS),” where I was able to learn firsthand the reasoning behind the new recommendations. The entire congress was themed “Oxygen – too little, too much, or just right?”, which of course made it the perfect setting to establish a new approach.

This recommendation now goes far beyond the maximum 210 minutes per 24 hours specified in the generally known NOAA table. How exactly did the scientists behind it come to throw these long-established rules overboard?
Let’s first take a look at what this table on O₂ exposure times actually is and where its weak points lie. To understand this, we first need to take a very basic look at the various forms of oxygen intoxication.

Main problem: CNS toxicity

The greatest concerns when using higher percentage decompression gases or very long exposures with a rebreather are known to be the risk of seizures caused by oxygen poisoning, which are almost impossible to survive underwater. The limit values used to try to reduce the risk to zero have been pushed further and further down for recreational use. At the moment, 1.4 bar pO₂ for the working phase of the dive, 1.6 bar pO₂ for the decompression phase and a setpoint of 1.3 bar pO₂ on the CCR are largely considered the norm, although there is a tendency to “prefer a little lower” (and of course there are still the “in the good old times we used to…” -guys).

There is a very good reason why these values are always under discussion and cannot be simply tested and established once. Unfortunately, the symptoms of CNS toxicity do not follow a simple dose-effect principle, but are above all one thing: chaotic and unpredictable. Data from the 1940s is still used today, in which experiments were carried out for the military in order to better predict the onset of oxygen poisoning. One of the most impressive results was that the same person on different days sometimes experienced symptoms after just a few minutes, but sometimes only after hours – under the same conditions! The tests were carried out at a pO₂ of around 3 bar (!) in water, but in a pressurized chamber, without further exertion at rest.
And the overview of the results from these studies also shows something important: nobody automatically gets symptoms at a certain moment, but even at oxygen partial pressures to which we would certainly not voluntarily expose ourselves in water, symptoms only occur after some time.

from: Groborz, O et al. (2025)New insights into the mechanisms and prevention of central nervous system oxygen toxicity: A prospective review. CC-BY 4.0
The data for the graphs are taken from Donald, K. W. (1947). Oxygen Poisoning in Man.

One would think that such “old” results are long outdated today – but the opposite is the case. These results were also shown at the SPUMS meeting in almost all presentations dealing with CNS toxicity. The reason for this is also obvious. Although the study is old, it has produced relevant results and measured the exposure at which symptoms of CNS toxicity occur in a significant number of test subjects. These symptoms are painful, unpleasant and not completely harmless. As the data is still good enough today, there is no reason to repeat the same series of tests and expose more people to it.

What becomes clear here, and should always be kept in mind, is that it is not possible to predict exactly when a particular person will develop symptoms. You can extrapolate certain probabilities from many test series and data from dives, but you cannot answer the question: How long can I use which pO₂ today without risking oxygen poisoning? Therefore, when limit values are set, the search is on for the limit below which the risk is as close to zero as possible.

Finding this limit is made more difficult by the fact that the risk is associated with factors that go beyond the pure oxygen partial pressure. In dry conditions, in a chamber, a higher pO₂ can be tolerated for significantly longer than under water. At rest, a CNS hit is significantly less likely than in motion, because CO₂ retention apparently also plays a role. With rebreathers there is an additional risk: the work of breathing is somewhat greater here than in an open system, and there is a possibility that the scrubber will no longer completely remove the CO₂ from the exhaled air.

Pulmonary toxicity

In addition to CNS toxicity, which is so important for us, there is something else that comes into play when assessing how long you want to be exposed to an elevated pO₂: pulmonary toxicity. Here the data situation is much simpler. Damage to the lungs almost inevitably occurs after longer exposure times, and this is measurable after longer pressurized chamber treatments. Here too, of course, there are individual differences, but the risk is much easier to assess. To prevent damage to the lungs, exposure times are controlled by OTUs (Oxygen Toxicity Units).
If the oxygen affects lung function, you will experience breathing difficulties, coughing, pain when inhaling – not nice, but in contrast to convulsions under water, anything but life-threatening. In most cases, the symptoms are reversible and the lungs recover.

The two forms of oxygen toxicity therefore have very different relevance when diving: while you want to avoid seizures at all costs because in most cases they will result in a fatality, you can accept the risk of reversible restrictions in lung function. Here you can weigh up whether the benefits of the higher pO₂ during decompression justify the risk.

The problem with the NOAA table

Oxygen exposure times according to NOAA (2001). Table 3.4, p. 3.23
The subdivision CNS vs Pulmonary Toxicity follows an idea by Simon Mitchell

There is also another problem: the limits are note only set, but the values are adjusted for shorter exposure times and repeated dives. To do this, percentages of the total exposure are taken and added together. This is exactly what is done in different forms in every O₂-CNS table and in every dive computer. The result is a model that is not validated in this form and is difficult to justify.

However, the frustration that the empirical basis here is so weak does not mean that the “CNS limits” do not work. They certainly do: accidents due to oxygen poisoning are extremely rare despite the worldwide spread of nitrox, and in the vast majority of cases are due to gases being mixed up. You can therefore continue to adhere to these limits and do nothing wrong.

The problem only arises when you undertake very long dives with very long decompression – rebreather divers, very experienced technical divers and underwater workers can get into this situation. In these areas, air breaks are used in an attempt to minimize the risk – but it is not uncommon for significantly longer exposure times to be accepted than is provided for in the table.

“To summarise the present situation, technical divers and other groups such as scientific divers conducting prolonged dives are working with restrictive oxygen exposure limits that are essentially untested for the outcome of greatest concern (CNS toxicity) in the range of inspired PO2s commonly used (< 1.6 atm). As technical diving has progressed to deeper longer dives, divers are inevitably forced to ignore these limits.” (Hoyt et al., 2025, p.231)

It has become common practice in this area to ignore the recommendations – which is a problem: the commonly used table is thrown to the wind. However, this is not always the result of an intensive examination of the topic and a conscious decision, but rather “that’s what everyone here does”. And this is where the warning lights should really be flashing red for every diver!

We now have a special situation. The limit values do not have a particularly good reputation, their scientific soundness has not been considered adequate for decades. And there are really good reasons for setting different limits. However, if this is not done in a consensual, well-thought-out way, but simply accepts the disregard of existing limit values, this can have an impact on the acceptance of these limit values as a whole, and can lead to a movement away from the rules that do work. This is precisely why it was so important to find a new, acceptable recommendation.

A new recommendation

The NOAA Workshop 2025 did not aim to throw the entire established oxygen exposure time table overboard. As Simon Mitchell put it: “Let’s not replace one crap evidence-free limit with another crap evidence-free limit”. Instead, the aim was to use the existing data to arrive at a recommendation that focuses on preventing CNS hits while not setting overly restrictive limits.
The limit of 1.3 bar pO₂ is based on the study situation. As already shown in the data collection by Vann et al. 2009, 1.3 bar pO₂ can be regarded as the limit at which the risk is close to zero. As the data situation is too weak for all higher pO₂ values, it was agreed to take this widely used pO₂ as the starting point.
After reviewing the data situation, the following consensus was reached:

With a constant (maximum) pO₂ of 1.3 bar, a total of 240 minutes diving time under normal load is acceptable. A further 240 minutes at rest during the decompression phase is also acceptable. This limit also applies if it is spread over several dives per day; the total time of 4 plus 4 hours should not be exceeded within 24 hours.
These limits are inevitably accompanied by an increased risk of pulmonary toxicity, which you must be prepared to accept. If symptoms occur, you should take a break from diving until they have completely disappeared.

Although the 1.3bar pO₂ can be considered so safe that a CNS hit is practically impossible, the authors recommend three strategies to minimize even the last risk.

Air breaks: It has been proven that short “air breaks” in pressure chamber treatments significantly reduce the risk of CNS poisoning. In the case of prolonged decompression, short breaks and switching to a gas with less oxygen or lowering the setpoint for five minutes are highly recommended.

Avoid hypercapnia: With greater exertion, the risk of CNS intoxication (and other discomforts) is significantly higher. Lower gas density and less exertion can help minimize this risk.

Fix the mouthpiece in the mouth: It can help not only with cramps, but also with other possible health problems if the mouthpiece is fixed in place.

Consequences of the new recommendation

This new guideline is a relief for those who occasionally exceed the previous CNS limits on their dives. They now have a well-founded new limit and can adapt their deco strategy accordingly.
Caution is advised if you would like to use 1.6 bar pO₂ in your decompression schedule as before. The new recommendations do not cover this for good reasons. For those who have already exhausted the “CNS clock” at 1.6 bar pO₂, further prolonged exposure to 1.3 bar pO₂ may entail unacceptably high risks. At the very least, the data situation is not clear enough for anyone to be prepared to make a recommendation that contradicts the established rules.

At the same time, the new recommendation does not mean that the limit must always be set at 1.3 bar pO₂. In the area of rather short exposures, which we have during the much more normal, much less extreme dives, we can of course continue to take advantage of a higher pO₂.

This article will appear in print in Wetnotes. This [German language] magazine is well worth a look for anyone interested in technical diving!

Thanks to the SPUMS for their support of this article. They sponsored access to their congress which was very relevant to this topic.