01The short answer
02What the numbers cover, and who they describe
A 2020 review reported 88.3 injuries per 1000 athlete-exposures in competitive sport karate. A 2019 analysis found that only about 10% of recorded injuries resulted in time lost from training.
Both describe elite competitors, under WKF competition rules, in tournament conditions. They are not measurements of knockdown karate and not measurements of ordinary class training. We give them because they are the best-quality figures available and because the alternative — no figures — is worse.
03Where the risk actually concentrates
The most useful finding in this literature is about distribution rather than rate. A 2018 study reported 43% of injuries occurring in sparring simulation, 26% in technical work, and 10% in competition.
That is the opposite of what most people assume, and it has a practical consequence: the thing that determines your risk is how your club runs its ordinary sparring, not whether you compete. Which is why how sparring is supervised is the question worth asking at the door.
04What gets injured
A 2019 study specific to Kyokushin (n = 61) reported contusions as 91% of injuries, with feet and hands the most affected sites.
A reservation we are not going to bury: that study appeared in a journal with weak editorial standards, and a single small sample is thin evidence. We report it because it is the only Kyokushin-specific injury data available, and we would rather show you its weakness than present it as more than it is. Clinical accounts from French-language sports medicine describe a broadly similar distribution.
Bruising being the dominant category is consistent with what the format does. It is not the same as saying nothing serious happens.
05Acute injury versus accumulated wear
The 2025 practitioner survey found that 30.1% of respondents had undergone surgery for their most serious injury and 26.2% reported training while injured. Age was not associated with injury prevalence (p = 0.221).
Self-reported, and drawn from people still training — the survivor bias is acknowledged by the authors. The 26.2% figure is the one worth sitting with: training through injury is common, and it is a cultural problem rather than a physical necessity.
06The head
A 2015 study (n = 24) found elevated biomarkers following head kicks in karate. That is a signal that something measurable happens, not evidence of lasting injury, and the sample is small.
What is established is the management protocol rather than the risk level: the 2023 Amsterdam consensus sets out a six-stage graduated return to play after a suspected concussion, and it is the standard any club should be following.
Knockdown rules exclude hand strikes to the face, which removes the highest-frequency source of head impact in striking sports. Head kicks remain legal. Those two facts together are the whole of the head-impact picture for this format — narrower than in most striking sports in one direction, and not zero in the other.
07What nobody has studied
There is no longitudinal cohort of karate practitioners. Nobody has followed a group over decades and recorded what happened to them, which means the question people most want answered — what does thirty years of this do to a body — has no evidential answer.
The cartilage study that occasionally gets cited covers nine athletes from unspecified disciplines and does not support conclusions about karate.
Absence of evidence here is neither proof of safety nor proof of danger. It is an honest gap, and anyone filling it confidently in either direction is going beyond what exists.
08What reduces risk, and is established
- Mouthguards. Training without one is associated with roughly 58% more dento-alveolar trauma. This is the clearest risk-reduction finding in the literature and the cheapest to act on.
- Weight categories. In competition, and in practice a matching principle for sparring pairs.
- Long warm-ups. Roughly 52% of injuries occur in the first minute of activity, which is an argument for the unglamorous part of the session.
- Matching by level. The single most controllable variable in ordinary training, and the one a club either does or does not do.
09The minor injuries you should expect
Setting expectations honestly is part of safety, because a beginner who does not know what is normal cannot tell when something is not.
Normal: bruising, particularly on the forearms and shins, in the weeks after partner work begins. General soreness in unfamiliar places. Sore feet, and the occasional stubbed toe, which is the most common minor injury in any barefoot art.
Not normal, and a reason to stop and ask: pain that is sharp rather than dull, pain localised to a joint, anything that is worse the next morning than it was that evening, and any impact to the head followed by feeling wrong in any way at all.
The distinction matters more than any figure on this page, because it is the one you will actually have to apply.
10Training while injured, which is the real problem
The 26.2% figure above is not about acute risk. It is about a culture in which stopping reads as weakness, and it is the mechanism by which a manageable injury becomes a lasting one.
It is a room norm rather than a requirement of the style, and it comes apart as easily as it forms: one senior belt saying out loud that he is sitting this one out because something hurts. A club where that sentence is awkward to say is telling you about its priorities.
This is also the part that is entirely within your control. The published data cannot tell you what thirty years of this does; it can tell you that a quarter of practitioners train hurt, which is a decision rather than an exposure.
11What this means for you
The risk is not primarily determined by the style. It is determined by how your club runs sparring, whether it matches partners sensibly, whether it warms up properly, and whether stopping is treated as normal.
All four are visible in one visit, which makes this the most actionable page in this section: you can assess your own risk profile before joining anything, by watching.
Keep reading
Ask how sparring is run before you ask anything else
The data says that is where the risk sits. It is a fair question at any club, and the answer should be specific.
Come and watch a session