A Not-So-Scientific Look at Injury in CrossFit

Recently researchers released one of the first scientific studies on injuries in CrossFit, but they decided to replace the laboratory with online CrossFit forums.

How often do athletes get injured doing CrossFit? This question evokes anecdotal evidence on both sides of the issue, but very little actual science. The authors of an upcoming study in the Journal of Strength and Conditioning Research noticed this lack of hard data. They decided to produce one of the first scientific studies on CrossFit injuries. Unfortunately, the methods they chose made the study only slightly more scientific than a Facebook poll.

Researchers posted a survey online to ten CrossFit discussion forums. The forums were chosen by Googling “CrossFit forum” and choosing the top ten search results. The survey was available for about three months, and the researchers received 132 anonymous responses. At this point, if you’re thinking, “This doesn’t sound like a very good way to perform a scientific study,” I’d have to agree.

Here’s why: What is the first thing an athlete does after sustaining an injury? He sits down at his computer to find out what the Internet has to say about the injury. I often tell my athletes to be careful when researching an injury online. Your injury could be a simple strain or sprain that will require a couple weeks of targeted rest. But no, we all focus on the worst possible outcome and follow the rabbit hole to its inevitable end: a torn labrum or cancer.

So the injured athlete goes online to get educated, and what does he find? A survey about injuries. His recent injury motivates him to take a look and respond to that survey. But what about the countless uninjured athletes who are electronically cruising past each second? They are far less motivated to respond. This is a classic example of sampling bias. The very nature of the survey attracted more injured athletes than non-injured ones. To be fair, the authors completely acknowledged this flaw. In an effort to mitigate it, they tried to encourage non-injured athletes to take the survey. But I’m just not sure that was enough.

The results can sound either terrible or great for CrossFit, depending on how you look at them. Over 73% of respondents reported an injury at some point during their CrossFit training, which averaged eighteen months. Wow, that sounds bad. But when compared to total hours trained, researchers found an average of 3.1 injuries per 1,000 hours of CrossFit training. The authors reported that this injury rate is actually similar to activities like general gym training, running, weightlifting, powerlifting, and gymnastics. So is CrossFit inherently prone to injury? No more than general gym training, according to this study. The study also found zero instances of rhabdomyolysis, despite the great Internet rhabdo-pocalypse of October 2013.

Now let’s be real. Do people get injured doing CrossFit? Yes. CrossFit athletes are often known for working hard and pushing their bodies to the limit. At the extreme bounds of athletic performance, injuries sometimes occur. But that’s not a condemnation of the program itself. Any strenuous physical activity can lead to injury. Weightlifting, running, skiing, tennis, auto repair – they can all end badly if you attack them with reckless abandon. But if you learn from a competent coach and perform in a supervised setting, then they are all relatively safe. The same is true for CrossFit.


1. Paul Hak, et al. The nature and prevalence of injury during CrossFit training. Journal of Strength & Conditioning Research. POST ACCEPTANCE, 22 November 2013. doi: 10.1519/JSC.0000000000000318.

Photo courtesy of CrossFit Impulse.

Leave a Comment

Do Not Sell My Personal Information