← All articles6 min read
By Dr. Sarah Okafor

Does Stretching Prevent Injury?

Illustration split in two: a runner doing a seated hamstring stretch on a track, and the same runner doing a dumbbell Romanian deadlift

Not by any amount the trials can measure. A meta-analysis of 25 trials and 26,610 people found stretching left injury risk essentially unchanged (risk ratio 0.963), while strength training cut it to less than a third (0.315). Stretching does make you more flexible, so keep it if you need the range. Just do not count it as your injury prevention.

The trial that should have settled it

In 2000, Pope and colleagues randomly assigned 1,538 male army recruits to two groups for 12 weeks of training. Both did an active warm-up before each session. One group also did a supervised 20-second static stretch for each of six major leg muscle groups, every time.

The stretch group recorded 158 lower-limb injuries and the control group 175. The hazard ratio was 0.95, with a 95% confidence interval from 0.77 to 1.18, which includes no effect. What did predict injury was fitness, measured with a shuttle run, along with age and enlistment date. The authors’ conclusion was that a typical stretching routine does not produce clinically meaningful reductions in injury, and that fitness may be an important risk factor you can change.

Recruits are not the general public, and six 20-second holds is one specific dose. So the reviews matter.

What the pooled evidence says

Lauersen and colleagues (British Journal of Sports Medicine, 2014) pooled 25 randomised trials with 3,464 injuries and split them by type of exercise. This is the most useful single table on the question, because it puts stretching beside the alternatives.

ExposureRisk ratio (95% CI)Reading
Stretching0.963 (0.846–1.095)No detectable effect
Multiple exposures combined0.655 (0.520–0.826)About a third fewer injuries
Proprioception / balance0.550 (0.347–0.869)About 45% fewer
Strength training0.315 (0.207–0.480)Less than a third of the injuries

A risk ratio of 1 means no change, and the percentages in the last column are our plain reading of each ratio. These rows come from different trials in different sports, so they are not a head-to-head race. The gap between stretching and everything else is still hard to ignore.

The result was not a one-off. A CDC-led systematic review (Thacker, 2004) found stretching was not significantly associated with fewer total injuries (odds ratio 0.93, CI 0.78 to 1.11). It stopped short of telling anyone to quit, saying there was not enough evidence to endorse or discontinue routine stretching.

One review leaves a door open

Behm and colleagues (2016) reviewed static, dynamic and PNF stretching in healthy active people. They also found static and PNF stretching had no clear effect on all-cause or overuse injuries, and that no data existed for dynamic stretching. Yet they still recommend stretching inside a warm-up that includes dynamic activity afterwards, citing reduced muscle injuries and better range of motion.

So a narrower benefit for muscle strains specifically has not been ruled out. All-cause injury, the number most people are asking about, is where the trials come up flat.

What stretching does do

There is also a finding that makes the strength-versus-stretching choice less of a choice. Afonso and colleagues (2021) compared the two head to head in randomised trials and found no difference in range-of-motion gains (effect size −0.22, CI −0.55 to 0.12). The studies varied a lot in design and population, and the authors asked for more research. On current evidence, though, strength training appears to compete with stretching on flexibility, and it has the far better injury number.

The cost of stretching right before you train

Measured immediately afterwards, the Behm review found performance changed by −3.7% after static stretching, −4.4% after PNF and +1.3% after dynamic stretching. Dose mattered: holds of 60 seconds or more per muscle group cost −4.6%, against −1.1% for shorter ones.

The practical catch is that most of those tests happened 3 to 5 minutes after stretching with nothing in between. When dynamic activity followed the stretch, no clear effect was observed. So a short stretch followed by a proper warm-up is unlikely to hurt you. Our guide to warm-up sets covers the part that comes next.

What to do with this

The recommendations below are our reading of the evidence above, not a finding of any one paper.

What this evidence cannot tell you

These papers are about injury prevention in sport and training, and the largest single trial here used army recruits. They say little about stretching for a diagnosed condition, where a physiotherapist may prescribe it for a specific reason. Do not drop prescribed rehab because of population averages. This summary rests on the published abstracts of the papers below, and it cannot tell you how much any of these numbers applies to your own training.

Sources

General information, not medical advice. If you have an injury or a medical condition, check with your doctor or physiotherapist before changing your routine.

Related

Frequently Asked Questions

Not by any measurable amount in the best trials. A 2014 meta-analysis in the British Journal of Sports Medicine pooled 25 trials and 26,610 people and found stretching had a relative risk of 0.963, with a confidence interval spanning no effect (0.846 to 1.095). Strength training, in the same analysis, came in at 0.315.
For soreness and injury risk, the timing does not seem to rescue it: a Cochrane review found no clinically important reduction in soreness from stretching before, after, or both. Stretch whenever it suits your goal. If you stretch right before lifting or sprinting, keep static holds under 60 seconds per muscle group and follow them with dynamic warm-up work.
Barely. The Cochrane review (Herbert 2011, 12 studies) found post-exercise stretching reduced soreness one day later by about 1 point on a 100-point scale. A large field trial of 2,377 people found stretching before and after reduced peak soreness by about 4 points, which the authors called very small.
Yes, and this is what stretching reliably does. A single session typically raises range of motion for under 30 minutes. Over weeks it increases the stretch you can tolerate, though a 2018 review of 26 studies found programmes of 3 to 8 weeks did not seem to change the muscle or tendon themselves.
Slightly and briefly, when measured right afterwards. A 2016 review put the average performance change at -3.7% for static stretching and -4.4% for PNF, against +1.3% for dynamic stretching. Holds of 60 seconds or more per muscle group cost more (-4.6%) than shorter ones (-1.1%). When dynamic activity followed the stretch, no clear effect was observed.
In the 2014 meta-analysis, strength training had the strongest estimate (relative risk 0.315), then proprioception or balance training (0.550), then programmes combining several exposures (0.655). These comparisons are across different trials, not head-to-head, so treat the order as a strong hint and not a precise ranking.