Skip to content

Is CrossFit dangerous? What the injury research says

CrossFit Wilanów

Search for “CrossFit injuries” and you get two kinds of results: videos of someone dropping a barbell on their neck, and gyms promising that their box is perfectly safe. Neither tells you much. The question worth asking is different: how many injuries happen per hour of training, how that compares with other sports, and what raises the risk. There is research on all three, and it can be summed up honestly.

About three injuries per 1,000 hours

The broadest review was published by Hülsmann and colleagues in the German Journal of Sports Medicine in 2021. They collected 32 studies covering 9,306 people who do CrossFit and pooled 14 of them in a meta-analysis. The result: 3.20 injuries per 1,000 hours of training (95% confidence interval 2.06 to 4.34). Put differently, if you train three times a week for an hour, 1,000 hours is more than six years.

A second number from the same paper: 30.3% of participants reported at least one injury during the period studied (between 14.9% and 56.1% in individual studies). That spread is not an accident. The studies define an injury differently (from “something hurt” to “I had to see a doctor”), ask about different periods, and almost all rely on questionnaires filled in from memory. The authors rated the certainty of the evidence as low for injury prevalence and very low for the rate per 1,000 hours. That does not make the numbers wrong. It means you read them as an order of magnitude, not as a measurement.

Individual studies cluster around that value. Hak and colleagues (2013) counted 3.1 injuries per 1,000 hours from 132 online survey responses. Montalvo and colleagues (2017) found 2.3 per 1,000 hours in 191 people from four boxes in Florida. Feito and colleagues (2018) surveyed 3,049 people and, depending on how they estimated hours trained, got between 0.27 and 0.74 injuries per 1,000 hours, with 30.5% of participants injured in the previous 12 months. That last paper shows how much the result depends on the way hours are counted.

Compared with other sports, CrossFit does not stand out

Keogh and Winwood (2017) brought together studies of the weight-training sports. Bodybuilding has the fewest injuries (0.24 to 1 per 1,000 hours), strongman (4.5 to 6.1) and Highland Games (7.5) the most. Hülsmann and colleagues place CrossFit in between, at the level of weightlifting and powerlifting. Klimek and colleagues (2018), who reviewed three comparative studies, came to a similar conclusion: a risk comparable to weightlifting, distance running, athletics, rugby or gymnastics, with the caveat that confidence is limited by the lack of randomisation and control groups.

Comparing with running needs care, because runners are studied differently: prospectively, with different injury definitions. In the meta-analysis by Videbæk and colleagues (2015), novice runners had 17.8 injuries per 1,000 hours and recreational runners 7.7. Do not compare these one to one with CrossFit. Keep the broader point instead: a sport many people think of as the “safe alternative” has its price too, highest for beginners.

Shoulders, spine, knees

The Hülsmann meta-analysis also gives a map of the body. The most commonly injured areas are the spine (26.8% of injuries), the shoulder (25.9%) and the knee (15.8%). Weisenthal and colleagues (2014) were more specific: the shoulder suffers most in gymnastic movements, the lower back in barbell lifts. In the Dutch study by Mehrab and colleagues (2017), 58.7% of injuries were overuse injuries rather than acute ones.

This matters in practice. Overhead movements and pull-ups load the shoulders, squats and deadlifts load the spine, deep squats load the knees. When a coach corrects the position of the bar overhead, or will not let you do half-range pull-ups, they are working on exactly the places the statistics point to.

What raises the risk, and what does not

This is where the evidence is weakest. Hülsmann and colleagues say plainly that findings on sex, training experience and previous injury contradict each other across studies. Keogh and Winwood found very few differences by age, sex or level. Two patterns, however, show up in several papers.

The first is the start. In Mehrab’s study, people who had trained for less than 6 months had 3.7 times the odds of injury of those with at least two years’ experience (odds ratio 3.69; 95% confidence interval 2.09 to 6.50). Feito noted a higher risk in the first year. That fits what is known from running: the body needs time to adapt to a new kind of load, and the technique of new movements does not settle in a week.

The second is the coach. In Weisenthal’s study, 14.6% of people whose coach was present and correcting technique throughout the session reported an injury, against 24.4% where the coach was there “most of the time” and 31.0% where they were there “sometimes” (p = 0.028). It is a survey, so it shows an association, not a cause. But the direction matches common sense and CrossFit’s Level 1 Training Guide, where the order is mechanics, then consistency, and only then intensity.

Training dose does not give a simple answer. In Montalvo’s study, injured athletes trained 7.3 hours a week on average against 4.9 for the uninjured, and those who competed were injured twice as often as those who did not (40% against 19%). Feito, on the other hand, found that per hour of training, people who trained fewer than 3 times a week had the higher risk. The studies did not test this, but the simplest explanation for both is the same: risk rises when volume or intensity jumps faster than your preparation. If you are starting out, there is no reason to train every day.

What it means for you

The risk of injury is real and in the same league as other strength sports. Not zero, not exceptional. Most depends on the first months and on whether someone is watching your technique as you add weight and volume.

On your side it comes down to a few decisions. Tell the coach about pain that feels different from tired muscles during the session, not after it. Do not chase the RX version before the movement is repeatable with a lighter weight. In the first month, do not keep adding training days, even when you feel like it. If you have had a shoulder, back or knee injury, mention it before your first class: those are exactly the places the coach will watch.

What we do at the box

There are at most 12 people in a class. The coach leads the group from warm-up to stretch and corrects technique along the way, rather than leaving anyone alone with a barbell. Every workout has a beginners’ version: we change the weight, the number of reps or the movement itself, so the workout keeps its purpose without overloading you. For your first class you arrive 15 minutes early, the coach asks about injuries and breaks in training and picks the version you start with. Under the club’s Terms you tell the coach before class about illnesses, medication and injuries that may affect training, and the Terms also recommend a medical check before your first class.

None of this brings the risk to zero. It brings it to a level where what you and the coach do in the room matters more than the statistics. If you want to see for yourself, start with one class and bring your questions.

Sources

Buy your first class

One class to find out whether we suit each other. No commitment. If you decide on OPEN on the day of your first class, we deduct the cost of that class.

Buy your first class
First class(opens in a new tab)