Research · Mobility & Flexibility

Mobility vs. Flexibility: What the Research Actually Says About Performance and Injury Risk

By Muhammad Dzulhisham · MSc Sports Coaching, NSCA-CSCS

"Work on your mobility" and "you need to stretch more" get used in gyms and physio clinics as if they mean the same thing. They don't, and the research treats them as genuinely different measurements — one of which keeps showing up as a weak, near-clinically-useless predictor of injury, and one of which keeps showing up as something closer to a real signal. The confusion isn't pedantic. It changes what actually belongs in a warm-up, what belongs in a dedicated flexibility session, and what a screening test is even measuring in the first place.

Two different measurements wearing the same word

A 2026 paper in Sports Medicine, co-authored by a group of the field's most active stretching and range-of-motion researchers, is built entirely around untangling this confusion.¹ Its argument is direct: range of motion (ROM) is the umbrella term, shaped by both modifiable factors — soft tissue extensibility and neural control among them — and non-modifiable ones, like bone and joint architecture. Flexibility, in the paper's framing, is only one of several trainable components sitting underneath that umbrella, referring specifically to the extensibility of the muscle and connective tissue around a joint. Stretching is one way to train it, not a synonym for it; resistance training performed through a long muscle length, and even foam rolling, produce comparable long-term gains in the same measurement.

"Mobility" doesn't appear in that paper as a defined technical term — it's coaching-floor language more than a research construct. But it maps closely onto something the literature does define precisely: active range of motion, the amount of a joint's motion an athlete can produce and control using their own muscles, versus passive range of motion, how far a joint can be moved by an external force — gravity, a partner, a strap — with no muscular effort involved. Passive range of motion is what a sit-and-reach test or a passive straight-leg raise measures. It is essentially flexibility. Active range of motion additionally requires strength through that range and the motor control to get there under load — which is closer to what most coaches actually mean when they say "mobility."

The practical gap between the two shows up constantly on a gym floor. An athlete can have a passive hip-flexion range well past 90 degrees on a table and still be unable to actively lift that same leg anywhere close to it in an unsupported position — the tissue is long enough, but the athlete lacks the strength and control to use the range that's available. Flexibility describes what's physically possible. Mobility describes what the athlete can actually do with it.

What actually predicts injury risk

If passive flexibility were the thing that mattered for injury, a simple hamstring flexibility test should reliably flag who's about to strain one. It doesn't. A prospective cohort study of 438 professional soccer players across two full seasons in Qatar's top division, tracking 78 hamstring injuries, found that passive knee-extension and ankle-dorsiflexion range of motion were statistically associated with injury risk — but the effect sizes were small (d < 0.2), and the actual gap between injured and uninjured players was only 1.8 degrees and 1.4 centimetres.² The paper's own title calls these findings what they are: weak risk factors, of little use for actually predicting which individual player gets hurt.

Where the picture gets more interesting is when researchers test active range of motion instead of purely passive length. A 2023 study of 100 amateur soccer players tested six separate hip and knee ROM measures against hamstring-strain history, split by sex.³ The strongest models, built from the right combination of tests, correctly classified roughly 95% of players in both sexes — a result nothing like the near-null findings for simple passive flexibility. Tellingly, the strongest single predictor in both men and women was the active straight-leg raise: not how far the hip could be passively pushed, but how far the athlete could actively lift and control the leg under their own muscular effort. The purely passive measures in the same test battery were far weaker predictors on their own.

The most direct head-to-head comparison comes from a six-month prospective study of physically active adults that measured both flexibility (lower back and hamstring) and the quality of fundamental movement patterns in the same subjects, then tracked who got injured.⁴ Movement-pattern quality predicted injury occurrence with 73% accuracy. Flexibility predicted it with only 41% accuracy. Athletes with poor movement quality were nearly seven times more likely to be injured; each centimetre of reduced flexibility raised injury risk by a comparatively modest 6%. Combining both measures didn't improve on movement quality alone. The authors' own conclusion is direct: "the quality of movement patterns is an accurate predictor of injury risk, but lower back and hamstrings flexibility is not a reliable predictor of injury."

Flexibility measures how far a joint can be pushed. Mobility measures whether an athlete can control what happens once they get there — and across three separate lines of research, control is the number that keeps predicting outcomes flexibility alone can't.

Does static stretching before sport help or hurt?

This is where decades of gym-class habit collide with a genuinely large body of acute-effects research. The foundational modern review — a 2011 paper in the European Journal of Applied Physiology that's been cited thousands of times since — established the pattern most subsequent studies have gone on to confirm: static stretching's effect on strength and power is dose-dependent, not fixed.⁵ Short holds are largely inconsequential; long, sustained pre-performance holds are where the problems start.

Two mechanisms are usually offered to explain why a long enough hold produces a real strength deficit. One is mechanical: stretching temporarily reduces the stiffness of the muscle-tendon unit, which can blunt how efficiently a tendon stores and releases elastic energy during an explosive movement. The other is neural: several studies have recorded reduced muscle electrical activity (EMG amplitude) after a sustained stretch, suggesting the nervous system is, at least temporarily, driving the muscle less hard. Which of the two dominates is still genuinely debated in the literature — some studies find changes in muscle stiffness that don't fully account for the force loss, and others find neural changes without the specific reflex signatures researchers expected. What's more settled than the mechanism is the practical shape of the effect: short holds don't reliably trigger it, and long ones do.

A 2024 systematic review with multilevel meta-analysis put numbers on exactly where that line sits.⁶ Across the pooled evidence, static stretching produced a small overall effect on maximal-strength loss (effect size –0.21). But that average hides a sharp threshold: for stretches held 60 seconds or longer, the effect size on strength jumped to –0.84 — a large, meaningful deficit. Below that duration, the strength cost was negligible. Critically, the same meta-analysis found stretching "did not negatively influence athletic performance in general" when the outcome was jumping, sprinting, or throwing rather than isolated maximal-strength testing — stretching even showed a small positive effect on subsequent jumping performance in adults. The authors' conclusion pushes directly against the blanket "never stretch before sport" advice that circulated for years: rigorously avoiding any stretching before performance is, in their words, "without evidence, especially since the literature shows that potential strength impairments can be counteracted by subsequent dynamic activities."

A 2025 Delphi consensus statement — a formal agreement process involving 20 international stretching researchers, requiring at least 80% agreement on each conclusion — reached a similarly nuanced landing spot rather than a blanket verdict.⁷ Stretching reliably improves range of motion, both immediately and with consistent training, and that's a genuine, well-supported benefit. But the panel also agreed stretching is largely ineffective as a post-exercise recovery tool, doesn't meaningfully lower overall injury risk (and may simply shift injury type from muscle strains toward bone and joint injuries rather than reducing injuries outright), and isn't the all-purpose posture-fixing, performance-boosting tool it's often marketed as.

The NSCA's own guidance draws a distinction worth keeping separate from all of the above: a warm-up and a stretching session are not the same activity wearing different names.⁸ A warm-up's job is to raise tissue temperature and rehearse the movement patterns about to be used; stretching's job is to improve flexibility. There is, per the NSCA, little scientific evidence that pre-event static stretching of the prime movers prevents injury or improves performance — and some evidence that stretching those muscles right before a power or speed effort can measurably work against it, particularly with longer holds.

What this actually means for timing a warm-up

• Short static holds — under roughly 60 seconds per muscle group — are not the villain older advice made them out to be, especially when they're followed by dynamic, sport-specific movement rather than sitting as the final thing an athlete does before competing.
• Long holds (60 seconds or more) right before a maximal-strength, sprint, or jump effort carry a real, dose-dependent cost. This is where the evidence is least ambiguous — save the long holds for a separate session.
• A dynamic warm-up and a flexibility session are different tools solving different problems. Collapsing them into one block right before training or competition is the most common practical mistake — not stretching itself.
• Improved flexibility is a legitimate, real training outcome. It's just not the same outcome as reduced injury risk or improved performance, and shouldn't be programmed as if it automatically produces either.

What this means in the gym

Put the two threads together and a coherent, if less tidy, picture emerges. Passive flexibility is measurable, trainable, and genuinely useful for certain goals — but on its own, it's a weak signal for who gets injured and a non-factor for most in-season performance outcomes. What keeps showing up as meaningful, across the hamstring-injury data and the movement-quality research alike, is whether an athlete can actively control force through the range they have — the strength and motor-control layer sitting on top of raw tissue length, which is what "mobility" is actually pointing at even where the research literature prefers more precise language for it.

That's the practical argument for training end-range strength and controlled movement quality directly, rather than assuming a pre-session static-stretch routine is doing that job by proxy. Loaded stretching (holding an external load at a lengthened joint position), eccentric work performed at long muscle lengths, and controlled articular rotations through a joint's full active range all ask the nervous system and the muscle to produce force in the exact position a passive stretch simply parks the athlete in. That's a meaningfully different training stimulus, and it's the one more consistent with what the injury-prediction research keeps flagging as relevant — control through range, not just the raw range itself.

It's also the argument for testing an athlete's actual movement quality and active control before writing a mobility program, instead of defaulting to a generic stretching routine because a joint "feels tight." Tightness, flexibility, and mobility are three different reports from the body. Tightness is a sensation, not a measurement. Flexibility is a passive-range number a goniometer or tape measure can produce in isolation. Mobility is a demonstrated ability to control force through that range, under load, at speed, in the position the sport actually asks for it. Programming a fix for one when the actual limiter is another is a common, avoidable way for a mobility routine to consume training time without changing anything that matters on the field.