What the evidence says about stretching

What people are told

That stretching lengthens muscles, prevents injury, reduces soreness, improves performance, and should be done before exercise.

That is five claims presented as one habit, and they have five different answers.

What is actually known

On range of movement: stretching works. Regular stretching reliably increases how far a joint can be moved. That much is well supported and not in dispute.

On why it works: less settled, and more interesting than the usual explanation. A substantial body of work supports stretch tolerance — that what changes is largely your nervous system's willingness to allow the position, rather than the muscle becoming permanently longer. The sensation of stretch is a nervous-system output; raise the threshold and range increases without the tissue being structurally different.

Some structural change appears possible with sustained, high-volume programmes. For ordinary stretching, tolerance explains most of the effect.

On injury prevention: no. Twenty-five randomised trials, 26,610 participants, no beneficial effect.¹ This is covered in warming up.

On soreness: minimal. Studies of stretching for delayed-onset soreness find small effects at best — see soreness.

On performance: static stretching immediately beforehand can slightly reduce force production. The effect is modest.

What follows from that

Stretch if you want more range. That is the goal it delivers, and it is a legitimate goal — some movements need range you may not have, and the goblet squat and pike push-up records both name range as the limiter.

Do not stretch expecting the other four outcomes. They are not there.

Loading through range does the same job and more. Strength training through a full range increases range as well, and it is the intervention with actual injury-prevention evidence behind it. If you are choosing one, that is the one.

Tightness is not the same as shortness. A muscle can feel tight because of how the nervous system is reporting it, and that sensation can have sources — including referred pain — that stretching will not address.

What we still cannot say

How much of the range gain is tolerance and how much is structural, for any individual or any protocol. The balance is argued about.

Whether the small performance decrement matters to anyone outside a competitive setting. Probably not.

Whether stretching helps anything that feels like it should be helped by stretching. The gap between how much better stretching feels and how little it demonstrably does is one of the wider gaps in this field, and "it feels good" remains a perfectly reasonable reason to do something.

Sources

1. Lauersen JB, Bertelsen DM, Andersen LB. The effectiveness of exercise interventions to prevent sports injuries: a systematic review and meta-analysis of randomised controlled trials. British Journal of Sports Medicine 2014; 48(11):871–877. doi:10.1136/bjsports-2013-092538 PMID 24100287 25 trials, 26,610 participants, 3,464 injuries. Stretching showed no beneficial effect (RR 0.963, 95% CI 0.846–1.095). Strength training reduced injuries to under a third (RR 0.315, 95% CI 0.207–0.480) and proprioception training also helped.

Every source here was checked against PubMed rather than recalled. Where this site says something is disputed or unsupported and no source appears, that is deliberate: it means the claim is being conceded rather than cited.