Why Divers Get DCS Within the Limits: Data and Risk Factors

How often does DCS happen within the limits?

Most of the time. DAN says the majority of decompression illness cases it manages occur in divers who stayed within the limits of their dive computer. A large dataset with recorded profiles comes from DAN Europe: 39,099 dives logged by 2,629 European divers over five years, plus 320 dives that ended in DCS. When the researchers ran every DCS dive through the unmodified Bühlmann ZHL-16C model, only 8 had exceeded its limit. They called the other 312 "undeserved".

The numbers depend on how "within limits" is defined, as the table shows. The DAN Europe study recomputed each dive with the base Bühlmann model, which is more permissive than a computer set with gradient factors below 100%. Another series judged each dive against the table the diver had used. Either way, a clean profile is no guarantee.

This page summarises published data and is not medical advice. Anyone with symptoms after a dive needs a medical evaluation, whatever the profile looked like. More on dive safety is in our safety guides.

Why a dive inside the limits can still cause DCS

Because a limit is a chosen level of risk, not a boundary between safe and bent. A 2024 review in Diving and Hyperbaric Medicine describes table and computer limits as a point on a continuum of risk that the designer considered acceptable. Designers choose different points. Richard Vann, interviewed by DAN, gave the US Navy's judgement as 2% acceptable for mild DCS and 0.1% for serious DCS, while DAN's Project Dive Exploration measured 0.01-0.04% in recreational diving. His summary: there is a finite probability of DCS for most dives, and when enough dives are made, some DCS becomes likely.

Risk also climbs well before the limit. DAN's Alert Diver calls DCS probabilistic, with risk rising the closer you get to the no-decompression limit (NDL), and says staying within 50% of the allowable bottom time carries a negligible risk. The DAN Europe data fit that picture. The average dive in the database reached 66% of the Bühlmann limit; the dives that led to DCS averaged 79%, and 73.7% of them fell between 70% and 90%. Our guide to no-decompression limits covers how those limits are built.

Then there is the diver. A US Navy analysis, summarised in the 2024 review, found that the same diver produced very different bubble levels after identical, tightly controlled dives, and that DCS almost always struck on the high-bubble occasions. A computer cannot see that variation. DAN's Neal Pollock adds that current computers do not account for temperature or exertion in any meaningful way, even models that measure water temperature or heart rate.

Dive-pattern risk factors: repeated days, fast ascents and flying

DAN names the dive profile (time, depth and ascent rate) as the most significant risk factor, and the ways a profile adds risk are not always visible on a single dive's screen.

• Staying near the limit, dive after dive. DAN's Nick Bird sees remaining close to NDLs throughout dives and dive series as a likely contributor to many cases. The 2024 review calls dive patterns provocative when they are technically compliant but close to the limits, especially with several dives a day on consecutive days.
• Multi-day series. In one DAN case summary, a diver on the third day of a two-dives-a-day resort series spent under a minute at 30 m (100 ft), about 10 minutes at 24 m and 15 minutes at 18 m, made a slow ascent and a 3-minute safety stop, and was diagnosed with serious DCS that needed several chamber treatments. DAN's comment was that it happened probably without violating the computer's limits.
• Fast ascents and missed stops. Both raise the risk, the 2024 review says, though breaking the rules does not guarantee DCS any more than following them prevents it.
• Flying or driving to altitude afterwards. Cabin pressure is lower than at sea level, so a flight adds decompression stress. The DAN/UHMS guidelines suggest at least 12 hours after a single no-stop dive and 18 hours after multiple dives or days, while the 2024 review recommends 24 hours in most recreational scenarios. Details are in our guide to flying after diving.

Body and behaviour: exertion, cold, PFO and dehydration

These are the factors a computer cannot measure, and the evidence behind them varies. Temperature and exertion have the strongest human data; dehydration has the weakest.

• Exertion at depth. Hard work on the bottom increases gas uptake. In the DAN Europe data, heavy exercise in the 24 hours before diving and heavy workload during the dive were both more common in the DCS cases.
• Exertion after the dive. DAN warns that exercise near the end of or soon after a dive, especially with high joint forces, can promote bubbles and help them bypass the lung filter. The DAN Europe study found venous bubbles peaking 30-45 minutes after surfacing, the window when many divers are still carrying gear.
• Temperature pattern. Being warm at depth and cold on the way up is the worst combination. The 2024 review cites human experiments in which that pattern carried at least ten times the risk of the reverse. DAN adds that a hot shower or hot tub straight after the dive can trigger skin bubbles.
• A large patent foramen ovale (PFO). This opening between the upper chambers of the heart is found in about 25% of adults, but the large ones linked to DCS are present in fewer than 2% of people. They let venous bubbles bypass the lungs and are associated with brain, spinal, inner ear and skin DCS, not with joint pain.
• Dehydration. DAN says it gets more attention than it deserves compared with the profile, temperature and exertion. The 2024 review finds human data largely lacking, though one small study showed that drinking about 1.3 litres in the hour before diving reduced bubbles.
• Sex, age and body fat. The DAN Europe data found women (41% of DCS cases against 17% of logged dives), older divers and divers with more body fat over-represented. DAN's own reference says there is little evidence that sex plays a role, so the sources disagree here.

How conservatism settings and gradient factors change your risk

They move your limit further from the model's edge, and that is the one lever your computer gives you directly. On a Bühlmann computer, GF High is the percentage of the M-value (the model's maximum tolerated supersaturation) that you may reach when you surface. Pollock explains in Alert Diver that on recreational dives GF High usually sets the greatest decompression stress of the dive, while GF Low is often never reached. Lowering GF High lowers that ceiling, which shows up as shorter no-stop times.

Pollock describes a setting of 30/70 as giving a greater buffer than 15/85, and his reason for having any buffer is that the M-value is a limit we know is not universally safe to approach. He also suggests lowering GF High during a dive if you worked harder at the bottom than expected. Other computers use their own conservatism levels instead, and DAN's Nick Bird points out that buddies on different computers or settings get different NDLs on the same dive.

The DAN Europe distribution shows why this matters. Most DCS dives peaked between 70% and 90% of the Bühlmann limit, and a GF High inside that band, such as 85 or 70, caps how far into it your dives can go. How gradient factors and brand presets work is covered in our guide to dive computer algorithms. Change settings only once you understand them, and set both computers in a buddy pair the same way if you can.

Practical steps that lower your risk

No step makes DCS impossible, and DAN notes that nothing short of not diving removes all risk. These are the measures the sources describe for keeping it low:

• Stay well inside your NDL rather than close to it. DAN's figure is that staying within 50% of the allowable bottom time carries negligible risk.
• Add margin for each risk factor. With cold water, exertion or deep dives, DAN's Nick Bird advises staying further from the NDL, making longer safety stops and considering longer surface intervals and fewer dives per day.
• Dive nitrox with air limits for a buffer. DAN says that using nitrox against air table limits reduces gas uptake. Pollock warns against doing it by setting the computer to air, because it then will not warn you about oxygen exposure.
• Stay comfortable but not hot on the way down, warm on the way up, and skip the hot shower or hot tub straight after the dive.
• Keep effort low on the bottom, and avoid heavy lifting and exercise after the dive. For divers with a known PFO, the 2024 review cites advice to avoid lifting and straining for at least three hours.
• Ascend at your computer's rate and never cut required stops.
• Before flying, wait at least the DAN/UHMS minimums of 12 hours after one dive or 18 hours after several dives or days; the 2024 review recommends 24 hours. DAN says a drive to altitude adds stress too.
• Set your computer's conservatism deliberately rather than leaving it at whatever it shipped with.

If you have symptoms after a clean dive

Treat them as possible DCS anyway. A hyperbaric physician writing in Alert Diver says a misunderstanding he consistently sees is divers insisting their DCS must have been random because they were within the limits. A clean profile does not rule DCS out.

Follow standard first aid (oxygen, local emergency services, then DAN) and keep the dive profile from your computer for the doctors. The full sequence is in our DCS symptoms and first aid guide. Afterwards, if the symptoms were of the kind linked to a shunt (brain, spinal cord, inner ear or skin), the 2024 review says testing for a PFO is advocated before more diving.