The short answer
For the kind of stretching most people mean (holding a position until a muscle feels long, then coming out of it) somewhere between twenty and sixty seconds per stretch is where the useful range sits. Below about fifteen seconds you're mostly just moving. Beyond a minute or two the extra return gets small enough that it's usually better spent on a different body part.
That is not just a rule of thumb. A 2025 meta-analysis in Sports Medicine pooled 189 studies covering 6,654 adults to work out the dose, and found that thirty-second holds produced the same gains as sixty-second ones. Longer was not better. It was just longer.
Why the numbers vary so much
Because "stretching" describes at least four different activities, and they have different clocks:
- A quick reset at a desk. Ten to twenty seconds. The goal isn't lasting change, it's interrupting a position you've held too long.
- A general mobility session. Twenty to forty-five seconds per stretch, which is what most of Lucy's routines use.
- Working on a genuinely restricted joint. Longer holds, repeated, over months. This is a project, not a session.
- Warming up before sport. Mostly movement rather than holds. Long static stretches immediately before explosive work can temporarily reduce power.
Most advice arguments online are two people describing different activities at each other.
Total time matters more than any single hold
The more useful number isn't how long one stretch lasts. It's how much time a muscle group gets across a week.
The same Sports Medicine review put numbers on both ends of that. Within a single session, the gains flattened at roughly four minutes per muscle group. Across a week, they flattened at roughly ten minutes per muscle group. And how those weekly minutes were divided made no difference: three sessions and six sessions produced the same result, as long as the weekly total was reached.
So five thirty-second holds spread over the week beat one long session, partly for that reason and mostly for a very human one: you'll do the short ones.
One more finding from that review is worth repeating, because it answers the objection most people actually have. The largest improvements came from the people who started out the least flexible.
This is why Lucy counts stretches in seconds rather than reps, and why every duration in a routine can be nudged up or down before you start. Thirty seconds is the default because it's long enough to be worth doing and short enough not to be a negotiation.
What the hold should feel like
A four or five out of ten. Present, unmistakable, slightly boring. Not something you have to breathe through.
Two signs you've gone too far:
- You're holding your breath. Discomfort you have to brace against is past the point of usefulness.
- The stretch has become sharp, or has moved into a joint. Stretch belongs in the meat of a muscle, not in a knee or a lower back.
If you're shaking, you're not stretching any more. You're doing an isometric hold at an awkward angle.
Both sides, roughly the same
When a stretch is one side at a time, do both, and give them similar time. It's tempting to spend longer on the tight side, and that's not wrong, but the difference should be one extra round, not five minutes versus thirty seconds.
Lucy handles this by counting each side separately and telling you when to switch, which mostly exists so you can shut your eyes and stop doing arithmetic about your own body.
The version to remember
Twenty to sixty seconds. Four to five out of ten. Breathe. Both sides. Come back in a couple of days.
If you want a number to hold onto: thirty seconds, most days, is a completely respectable stretching practice, and it beats the perfect one you're planning to start next month.
Lucy is a stretching app, not a clinician. If a stretch causes sharp pain, numbness or anything that travels down a limb, stop. Then read our Fitness & Safety Disclaimer.
FAQ
How long should you hold a stretch?
Twenty to sixty seconds covers it, and thirty is a sensible default. The 2025 Sports Medicine meta-analysis of 189 studies found thirty-second holds producing the same gains as sixty-second ones, so there is nothing to be won by grinding through a longer hold.
Is it better to hold a stretch longer or repeat it?
Repeat it, or spread the time across more sessions. That review found the weekly total per muscle group to be the number that mattered, with no difference between three sessions a week and six as long as the total was similar.
How many seconds is too long to hold a stretch?
There is no harmful upper limit for a gentle stretch, but the returns flatten: in that review, gains within a single session levelled off at around four minutes per muscle group. Beyond that, the time is better spent on a different area.
Should you hold a stretch until it stops feeling tight?
Not as a rule, and that sensation varies too much day to day to use as a timer. A count is more reliable. If you are shaking or holding your breath, you have gone past a stretch into something else.
Does holding a stretch longer before sport help?
Long static holds immediately before explosive work can temporarily reduce power, so movement is the better warm-up. Save the holds for afterwards, which is what stretches after a run covers.
Lucy holds the timing for you, so you never have to count a thirty.