Injuries in BJJ: what the data actually says
Per hour spent on the mat, competition is ten times riskier than training. And yet nine injuries out of ten happen in training, and the reason is simple.
BJJ Square · Updated on 08/11/2026
Careful This section describes what research has measured, not what you should do. It is not medical advice: for your own case you need a professional who can examine you. Where the data is missing, we say so.
There are by now several studies on jiu-jitsu, and put together they tell a fairly consistent story — with one large caveat that is worth stating up front.
⚠️ Why the numbers seem to contradict each other
Two studies of the same events can report injury rates that differ by a factor of four. Not because one of them is wrong: because the definition of injury changes. A Hawaii study with orthopaedic surgeons at the mat side counted 9.2 injuries per 1,000 entries; a study at the no-gi World Championship, with broader criteria, 38.6. Before quoting a number you need to know what was counted.
Competition and training: the paradox
Per unit of time, competition is between eight and ten times riskier than training. But training is where almost all the time is spent, and that is where the vast majority of injuries occur: a study of 1,140 practitioners places 77.6% in sparring and only 9.6% in competition; another, of 881 people, reports 89% in training.
Note Put usefully: if you train and never compete, you are not avoiding the risk, you are only diluting it. And sparring — not technique, not the warm-up — is when almost everything happens.
What gets injured, and where
- In training the knee leads, at around 25% of injuries, followed by the shoulder at 13-15%. Three independent studies, of 341, 881 and 1,140 people, land on almost the same figure.
- In competition the elbow leads: 38.9% in a study of over 5,000 entries, and in ten cases out of fourteen the cause was an arm lock.
- Hands and fingers come out as the single most affected area in one study of training injuries (14.4%), but almost nobody takes them to a doctor.
The figure worth remembering
Among injuries caused by a submission, 78.9% land on the person receiving it, not the one applying it. The most heavily represented techniques are the armbar (22.4%), the kimura (12.6%) and the heel hook (11%).
Careful ⚠️ Be careful about what this number says and does not say. The study measured who got hurt, not why: it does not separate the people who tapped late, the people who caught a submission applied too fast, and the people who tapped in time when the damage was already done. So you cannot say that four injuries in five are the fault of someone not tapping.
We cover it here: tapping early.
How serious they are
Here the data is shakier. A British study of 372 injuries classifies 49.5% as “major”, meaning more than 28 days out. It is a high figure and it needs handling with care: these are self-reported recollections, and serious injuries are remembered while minor ones are forgotten. Another study, of a single club of 70 people, reports 15% of injuries requiring surgery — small sample, one city.
⚠️ What research does not know
- There is no prospective study on BJJ: everything is retrospective and self-reported, and recall accuracy at twelve months is estimated at around 61%.
- There is no female injury profile: every study has samples that are 88-90% male.
- No preventive measure has shown an effect: one study tested sleep, diet, mobility, strength and hot-and-cold therapies without finding significant associations. ⚠️ That does not mean they are useless: it means that study design could not demonstrate it.
The main sources for this page are studies published in the *Orthopaedic Journal of Sports Medicine*, *Journal of Athletic Training*, *Sports Health* and *BMJ Open Sport & Exercise Medicine* between 2014 and 2026.
Frequently asked questions
- Is BJJ a dangerous sport?
- Per unit of time, competition is clearly riskier than training, but the damage profile is joint-related rather than craniofacial as in striking sports. Reported rates vary a lot between studies because the definition of injury changes.
- What is the most common injury in jiu-jitsu?
- In training, the knee, at around 25% of cases, followed by the shoulder. In competition the elbow leads, almost always from an arm lock.
- Do people get injured more in competition or in training?
- In competition the risk per unit of time is eight to ten times higher, but around nine injuries out of ten still happen in training, because that is where almost all the time is spent. Sparring is the most represented moment.
- How many injuries require surgery?
- One study of a single club of 70 practitioners reports 15%, but the sample is far too small to generalise from. There are no sounder estimates.
Share this guide
Read next
- Tapping in time: the only skill that genuinely prevents anythingAmong injuries caused by a submission, 78.9% land on the person receiving it. It is the most useful figure jiu-jitsu research has produced.
- Ringworm, herpes and impetigo: skin infections in grapplingThe subject nobody enjoys bringing up, and the one that gets you turned away at a competition skin check. What has been measured, what is just custom, and what nobody has studied.
- Training after 40 without falling apart: what we know and what we don'tPeople training after forty want precise answers and find opinions instead. The reason is that studies on this age group in grappling simply do not exist.