By Raji Shah ยท Updated 2026-10-24
Running power puts a number in watts on your effort, borrowing the idea from cycling. But unlike a bike's power meter, running power is an estimate. Every brand calculates it differently, the numbers from different devices don't match, and there is little evidence that training by power makes you faster. Used consistently on one device, it can still be a useful gauge of steady effort, especially on hills.
What running power is, and isn't
On a bike, a power meter measures the force you put through the pedals and how fast they turn. Running has nothing equivalent to measure. Most of the work your muscles do in running stays inside the body: one review put it bluntly, "unlike cycling, running entails negligible external mechanical work," and tendons and other elastic tissues supply up to half of the power each step needs. So running power is a model, estimated from your speed, the gradient, your body weight and, depending on the device, how your foot or wrist moves and the wind.
The brands don't even agree on what they are estimating. Garmin describes its figure as the power a runner applies to the road, Polar as a change in the mechanical energy of the body, and Stryd as "a direct reflection of a person's metabolic effort." Polar's own documentation says "there is no universally accepted definition for running power."
How each brand estimates it
| Brand | What you need | Inputs, as the maker describes them | Wind |
|---|---|---|---|
| Garmin | Many current watches from the wrist, or a compatible running-dynamics accessory | Running dynamics, body mass, environmental and sensor data | An optional "Account for Wind" setting, using the watch's speed, heading and barometer plus wind data from your phone |
| Apple Watch | SE, or Series 6 and later, in an Outdoor Run | Apple doesn't publish them | Not stated |
| COROS | All COROS watches from the wrist, outdoors only; a Stryd pod indoors | Weight, speed and other metrics, with more power uphill at the same pace | Not mentioned |
| Polar | Wrist-based on the Grit X range, Pacer Pro, Street X, Vantage M3 and Vantage V range; a sensor on the Pacer and Vantage M and M2 | Speed and gradient from GPS and the barometer, plus body weight | Not mentioned |
| Stryd | A foot pod | The foot's motion in three dimensions, plus impacts and forces | Measured by the pod; Stryd says air typically adds 0 to 10 W outdoors |
How accurate is running power?
Different devices give different numbers. In a study of 11 runners on a force-plate treadmill, a Stryd pod, a Garmin chest strap and a treadmill's own system all rose in step with speed, but "absolute running power measurements were different among devices." In a small 2024 study of five trail runners, Garmin power from a chest strap read 82 W higher than Stryd on average over the same sessions, and 109 W higher uphill, even though two units of the same brand agreed closely with each other. Stryd says its numbers and Garmin's "are not comparable and are not relatable to one another."
Repeatability depends on the device. In a study of 12 trained male runners comparing five systems, the Stryd pod was the most repeatable, with day-to-day variation of about 4% or less, and the most closely related to oxygen use. The Polar and Garmin systems of that era showed low repeatability. A 2023 study found Stryd readings at the same steady effort on different days within about 8 W of each other.
It tracks effort, mostly. In most lab studies, running power rose in line with oxygen use, with correlations high enough to be useful. One study of 24 runners, using an early chest-strap version of Stryd, found only a weak link, and another found Stryd tracked mechanical power closely while underestimating its absolute value by 38 to 60%. There is little peer-reviewed validation yet of wrist-only power from Garmin, Apple Watch or COROS.
It misses changes in efficiency. Because power is calculated from movement, it doesn't see how economically you run. In one study, at the same speeds, metabolic power was about 5% lower in a carbon-plated racing shoe than in a basic trainer, but Stryd power didn't change. In another, changing stride and arm use changed metabolic cost without changing power. And in a third study, after 20 weeks of training, runners were using more oxygen at the same power readings, so the link between their power and their effort had shifted.
Where running power helps
Hills. This is power's strongest case. In trained runners on a treadmill, power at the lactate threshold was the same on 0%, 5% and 10% gradients, and holding a fixed power on gradients up to 8% produced the same heart rate, perceived effort and breathing, although power underestimated the cost of the 6% and 8% grades by about 2 to 4%. Most of this evidence comes from treadmill tests at moderate gradients, and it's thinner for steep trails and for downhill running. Holding a steady power up a hill is still a more consistent guide than trying to hold your flat pace; our guide to pacing hilly races covers the effort side.
Fast changes in effort. Heart rate takes a minute or more to catch up with a change in speed, which makes it a poor guide at the start of a rep. Polar says power responds to an increase in intensity almost instantly, which is a manufacturer claim rather than a measured comparison. Our heart-rate zones guide explains where heart rate works better.
Threshold and critical power. In trained runners, the power produced at critical speed, roughly the boundary between hard and very hard running, matched critical power calculated from time trials, and one study on a track found power modeled performance better than speed from a GPS watch or foot pod. But track and treadmill numbers don't transfer: critical power on a track was at least 89 W higher than on a treadmill in one study, and most of the research establishing that boundary in runners has used critical speed rather than power.
Does training by power make you faster?
There's no evidence yet that it does. A 2026 review of 29 studies of the Stryd pod found that all of them measured runners at a single point in time rather than over a training period, and called for longer-term trials. Running power is a pacing and feedback tool. It doesn't replace a good training plan, and its clearest advantages over pace, heart rate and perceived effort are on hills and in fast changes of effort.
How to use running power
- Stick to one source. Don't mix wrist power, a chest strap and a foot pod, or compare numbers across brands. Pick one and compare it only with itself.
- Set your own zones. Garmin says its default power zones are based on gender, weight and average ability "and may not match your personal abilities." Use a threshold or critical-power test from your own device instead.
- Keep your weight setting current, because every model uses body mass in the calculation.
- Use it on hills and rolling courses to hold an even effort, and for the start of intervals, where heart rate lags.
- Retest indoors and out separately. Treadmill power and outdoor power don't match, and COROS's wrist power doesn't work indoors at all.
- Expect racing shoes to change the numbers. In one study, runners in carbon-plated racing shoes produced about 6% more power for the same metabolic cost as in basic trainers, so a race target set in trainers may undersell what you can hold.
- Check it against feel. If power says easy and your legs say otherwise, trust your legs; our guide to running by feel explains why.
Who should bother
Power is most useful for trail and hill runners, triathletes who already train by power on the bike, and runners who want a steady-effort number that reacts faster than heart rate. For most road runners, especially newer ones, pace, heart rate and perceived effort already cover what power offers. If your watch shows power for free, try it on a hilly route and see whether it helps you pace; there's no need to buy extra hardware for it.
Bottom line
Running power is an estimate of effort, not a measurement like cycling power. Its numbers differ between brands, it can't see changes in running economy, and there's no proof that training by it makes you faster. Within one good device, though, it's a responsive and consistent gauge of effort on hills and through fast changes in pace, and that's where it earns a place on your watch screen.
FAQ
Is running power accurate?
It's consistent within a good device more than it's accurate in absolute terms. Studies found the Stryd pod repeatable from day to day and closely related to oxygen use, but different devices give very different numbers for the same run, in one study about 80 W apart.
Can I compare my running power with someone else's?
Not meaningfully, unless you use the same device and model, and even then body weight and running style affect the number. Power is for comparing yourself with yourself.
Do I need a Stryd pod for running power?
No. Many Garmin, COROS and Polar watches and Apple Watch SE and Series 6 and later estimate power from the wrist. A pod adds Stryd's own wind sensor and works indoors, where COROS wrist power and Apple Watch power don't, and it was the most repeatable device in a five-way comparison.
Is running power better than heart rate?
For some jobs. Power responds quickly to changes in effort and handles hills well, while heart rate lags by a minute or more and drifts on long or hot runs. Heart rate still tells you how your body is coping, which power can't.
Does running power account for wind?
Stryd says its pod measures wind, though one study says it's still unclear how well its power accounts for air resistance. Garmin offers an optional setting that estimates it from the watch's motion and barometer and your phone's weather data. Apple, COROS and Polar don't mention wind in their documentation.
Will training by power make me faster?
There's no evidence yet that it does. A 2026 review of 29 studies found no long-term training trials. Use power as a pacing tool, especially on hills, not as a training method in itself.