Can you hold your FTP for an hour? What the research actually shows

Where the one-hour idea came from

The hour is a definition of convenience, not a measurement anyone took on you. Coggan and Allen defined FTP as the highest power sustainable in a quasi-steady state for approximately one hour [Allen et al. 2019]. Almost no rider's FTP was ever established by riding for an hour — it came from a 20-minute effort multiplied by 0.95.

The definition was pointing at a physiological boundary, not a stopwatch. FTP was meant to mark the divide between the heavy and severe intensity domains — the wattage below which lactate and oxygen uptake settle, and above which they climb until you stop. That boundary is a real construct, though in the literature it is critical power that marks it, not FTP [Poole et al. 2016]. The hour was shorthand for a long time near that boundary, chosen because it is a duration cyclists understand. It was never the thing being measured.

The number you actually own came from a protocol. The 20-minute test takes 95% of your best 20-minute average [Allen et al. 2019]. TrainerRoad's ramp test takes 75% of peak one-minute power. TrainingPeaks watches for a new 20-minute best and suggests a threshold increase at the same 0.95, which you accept or ignore. Zwift and TrainerRoad both ship ramp and 20-minute variants. Not one of these protocols involves riding for 60 minutes, so testing the result against a 60-minute effort holds the number to a claim it never made.

The boundary itself is real; the duration bolted onto it was always loose. Poole and colleagues note that critical power was originally defined as the output sustainable indefinitely, or for a very long time without fatigue, then state plainly that this definition should be considered theoretical, since no exercise can ever be undertaken indefinitely [Poole et al. 2016]. What the boundary separates is power outputs for which exercise tolerance is predictably limited from those for which it is not.

What time-to-exhaustion at FTP actually measures

Two studies rode cyclists to failure at their own FTP. Wong and colleagues measured 33.7 plus or minus 7.6 minutes in 13 cyclists [Wong et al. 2022]. Borszcz and colleagues measured 50.9 plus or minus 15.7 minutes in 23 trained cyclists [Borszcz et al. 2018]. Neither mean reaches 60. Blowing up at 35 minutes puts you inside the published distribution.

The Wong result is the sharper one. Thirteen cyclists rode to task failure at FTP, then again at FTP plus 15 watts. Tolerance fell from 33.7 minutes to 22.0 plus or minus 5.7 minutes — a 15-watt step cut the sustainable duration by a third. End-test blood lactate differed significantly between the two intensities, roughly 6.7 against 9.2 millimolar. The authors concluded that FTP should not be considered a threshold marker between the heavy and severe intensity domains [Wong et al. 2022].

The gap between 33.7 and 50.9 minutes is not a contradiction to resolve — it is the finding. Different cohorts, different FTP determination protocols and different laboratories produced means 17 minutes apart. If time-to-exhaustion at FTP were a stable property of the metric, two studies of trained cyclists would not land that far from each other. Both figures are group means with wide standard deviations, which describes a population rather than predicting any individual inside it.

The rest of the threshold literature sits in the same band. Jones and colleagues, reviewing critical power as the maximal metabolic steady state, describe exercise at CP as sustainable for no more than approximately 20 to 30 minutes [Jones et al. 2019]. Poole and colleagues put the ceiling near 30 minutes for efforts above the boundary [Poole et al. 2016]. Whether you anchor to FTP or to CP, measured tolerance clusters between 20 and 50 minutes. The hour sits outside the range in every dataset here.

Why hold-duration is individual and your FTP number doesn't contain it

Two riders with identical 300-watt FTPs can differ by 15 minutes or more in how long they hold it. The standard deviations say so: plus or minus 7.6 minutes in one cohort [Wong et al. 2022], plus or minus 15.7 in the other [Borszcz et al. 2018]. Duration is a second, independent property, and your FTP does not encode it.

The mechanism is where your estimate lands relative to your true boundary. If your FTP sits a few watts below your critical power, you are riding in the heavy domain and tolerance is long. If it sits a few watts above, you are in the severe domain from the first pedal stroke, spending a finite anaerobic work capacity that empties on a predictable schedule [Poole et al. 2016]. McGrath and colleagues showed CP and FTP are not interchangeable in highly trained athletes, and the offset between them is rider-specific [McGrath et al. 2021]. Same 300 watts on the screen, two different intensity domains.

Measurement error stacks on top of that. Borszcz and colleagues compared FTP derived from a 20-minute test against mean power from an actual 60-minute time trial in 23 cyclists: the bias was only 1 to 5 watts, but the limits of agreement ran from 40 watts under to 32 watts over [Borszcz et al. 2018]. That is the difference between group agreement and individual prediction. The two methods agree on average and disagree on you, which is why the authors concluded they should not be used interchangeably on an individual basis.

We should be straight about our own limits. AdaptCycling does not estimate your individual time-to-exhaustion at FTP, and we do not prescribe against one. We fit a critical-power model to the power curve in your Strava history to estimate threshold, and that pipeline outputs a wattage, not a duration. Some platforms do publish a duration: TrainingPeaks lists Time to Exhaustion alongside its modelled FTP, defined as the predicted time an athlete can hold that power, and Xert ships a live time-to-exhaustion field. Intervals.icu reports eFTP, and Garmin's FTP estimate needs a paired power meter as well as a heart-rate strap. So this gap is ours, not the industry's — and where a platform does hand you a clock, it is a model's prediction, not a measurement of you.

A better validity check than the hour test

Stop using can I hold it for an hour as a validity test. Judge the number by whether prescriptions built on it land correctly — threshold work should feel like threshold work, and sweet spot should be finishable. That check costs one session. An hour-long time trial costs several days of recovery and settles nothing about the estimate.

The functional test is 2x20 minutes at 95 to 100% of your FTP with 5 minutes of easy spinning between. If you complete both intervals at the same power, with the second hard but controlled, the number is working. If rep two disintegrates before minute 12, the estimate is high — most often because a ramp test flattered a large anaerobic capacity. If you finish comfortably and could plausibly have started a third, the estimate is low. That is a decision you can act on inside a single Tuesday.

Cross-check against your own data rather than a fresh test. This is the broader argument our ftp-without-a-test pillar makes — estimating threshold power from the rides you have already done, without a formal test — and it carries more force once you accept that an hour-long time trial validates nothing. Ninety days of riding already holds dozens of near-maximal efforts across durations. If your modeled FTP and your best real 20 to 40 minute efforts disagree by more than about 5%, one of them is wrong and that is worth investigating.

For pacing a long effort, treat the hour question as an open one about you specifically. If a 40-kilometre time trial or a 50-minute climb is the goal, ride the first attempt 5 to 10 watts under FTP and record where you actually came apart. Do that twice and you own a personal number no model gave you. The failure mode to retire is not a wrong FTP — it is the assumption that the number arrived with a duration attached.

Quick answers

Sources cited in this guide

1. 01
   Wong et al. 2022. Functional threshold power is not a valid marker of the maximal metabolic steady state. Journal of Sports Sciences.
2. 02
   Borszcz et al. 2018. Functional Threshold Power in Cyclists: Validity of the Concept and Physiological Responses. International Journal of Sports Medicine.
3. 03
   Jones et al. 2019. The maximal metabolic steady state: redefining the gold standard. Physiological Reports.
4. 04
   Poole et al. 2016. Critical Power: An Important Fatigue Threshold in Exercise Physiology. Medicine & Science in Sports & Exercise.
5. 05
   McGrath et al. 2021. Do Critical and Functional Threshold Powers Equate in Highly-Trained Athletes?. International Journal of Exercise Science.
6. 06
   Allen et al. 2019. Training and Racing with a Power Meter (3rd ed.). VeloPress.

More inside FTP without a test

Other articles in this series

1. 01

   Indoor vs outdoor FTP: why the numbers differ

   Why your indoor FTP reads lower than outdoor — heat, cooling, motivation, and power-source differences — and whether to keep two numbers.

   →
2. 02

   How to estimate FTP without a power meter

   Estimating FTP from heart rate, RPE, and Strava when you don't own a power meter — how close you can get and where the method breaks down.

   →
3. 03

   20-minute vs 8-minute FTP test: which to use

   How the 20-minute and 8-minute FTP tests differ, the multipliers each uses, and which one fits your riding — plus why both are only protocols.

   →
4. 04

   Why cycling apps show you different FTP numbers

   Strava, Xert, Intervals.icu, and TrainingPeaks can each report a different FTP. Why the estimates diverge and which one to trust.

   →
5. 05

   Is your ramp test FTP too high? Why it happens

   Ramp tests overestimate FTP for anaerobically-gifted riders and underestimate it for diesels. Why the 75% rule misfires and how to correct it.

   →
6. 06

   How often should you test your FTP?

   How often to re-test FTP as a self-coached cyclist — twice a year, not every six weeks — and why modeled estimates change the cadence.

   →
7. 07

   Critical Power vs FTP: which number should you trust?

   CP and FTP track each other across a group but can differ 20-36W for you specifically. Why there is no correction factor, and which to anchor to.

   →
8. 08

   FTP stopped improving: are you a non-responder?

   Non-responder labels flip for 45% of people depending only on the statistics used. What a flat FTP usually means instead, and what to check first.

   →
9. 09

   Durability: why your power fades late in long rides

   What durability is, why the kJ threshold your app uses is arbitrary, and the honest evidence on whether it is independent of FTP or trainable.

   →