What the test factors are doing
Functional threshold power is defined loosely as the highest power a rider can hold in a quasi-steady state for around an hour. Almost nobody tests that way, because an hour at maximum is miserable to execute and worse to pace, so every practical protocol measures something shorter and scales it down.
The 20-minute test takes 95 percent of the average because a fit rider can typically hold about 5 percent more for twenty minutes than for sixty. The ramp test takes 75 percent of the best one-minute power at the end of a rising staircase. The two-by-eight protocol takes 90 percent. These multipliers are population averages fitted to groups of riders, and an individual can sit several percent either side of them.
That matters because the multiplier is applied before anything else. A rider whose true ratio is 0.92 rather than 0.95 walks away with an FTP roughly 3 percent high, and every zone above it is 3 percent high too, which is enough to turn a threshold session into an over-threshold session that cannot be completed as written. The adjustment field exists for exactly this: if your threshold intervals consistently fall apart in the last five minutes, the number is too high, not your willpower.
Which protocol suits which rider
| Protocol | Factor | Suits | Fails when |
|---|---|---|---|
| 20 minutes | 0.95 | Riders who can pace an even effort | You go out too hard and fade, or you sandbag the first ten minutes |
| Ramp to failure | 0.75 | Riders new to testing; short and self-pacing | You have a large anaerobic contribution, which inflates the result |
| Two 8-minute efforts | 0.90 | Riders who prefer shorter maximal blocks | The second effort is much weaker than the first |
| A hard race or long climb | 0.98 | Anyone with real 40 to 60 minute data | The effort was surging rather than steady |
The protocol you pick matters less than using the same one every time. FTP is most useful as a tracked series, and a series only means something if the measurement method is constant. Switching protocols mid-season produces a step change in the number that has nothing to do with fitness.
What a zone is and, more usefully, what it is not
A zone is a bucket. It exists so that a coach can write "ninety minutes in zone 2" and have that mean roughly the same thing next week as it did last week. The physiology underneath is continuous: there is no line in the body at 76 percent of FTP where tempo begins. Two widely used systems draw the boundaries in different places and both work, which is the clearest evidence that the exact numbers are conventions.
A zone is also not a measure of how hard something feels on a given day. Power is the work you are producing, and it is blind to heat, altitude, sleep, illness and how many days in a row you have ridden. The same 250 watts is a comfortable tempo in March and an unmanageable effort on the fourth day of a heatwave. Riders who treat the number as an instruction rather than a description end up training through fatigue that should have stopped them.
And a zone is not a health measurement. Nothing here describes a safe intensity, a target heart rate, or a threshold that applies to your body. It is a scheduling framework for a bicycle. If you have a cardiac condition, take medication affecting heart rate or blood pressure, or are returning from illness, the question of how hard to ride belongs with a clinician before it belongs with a spreadsheet.
Watts per kilogram, and its limits
Dividing FTP by body mass gives the number people quote, and on a sustained climb it genuinely predicts a great deal, because gravity is the dominant force and gravity scales with mass. The climbing time and power calculator shows that split directly: on an eight percent grade the gravity term is around ninety percent of the total.
On flat ground it predicts almost nothing. There, the dominant force is aerodynamic drag, which scales with frontal area rather than mass, and a heavier rider carries more absolute power against a frontal area that grew far less than their weight did. This is why the same rider can be dropped on every climb and comfortable in every flat rotation, and why comparing W/kg between riders who ride different terrain is close to meaningless.
The second figure this calculator shows, including bike and kit weight, is the one that actually governs a climb, because the bike goes up the hill too. On a nine kilogram bike with bottles, an 80 kilogram rider carries roughly eleven percent more mass than the headline number suggests.
Where the zones go next
Zones describe intensity, not volume, and volume is what most amateur weeks are actually limited by. Once the bands are set, the next questions are how long the sessions are and how much fluid the hard ones cost. Duration and pacing across other endurance sports are covered by the training pace zones calculator, and the fluid side by the sweat rate and hydration calculator, which is worth running for any threshold session longer than an hour in warm weather.
If you also train by heart rate, the two systems are not interchangeable. Heart rate lags power by tens of seconds, drifts upward across a long effort at constant power, and moves with heat and hydration. The heart rate zones calculator builds those bands separately, and where the two disagree during a session, power describes what you did and heart rate describes what it cost you.
Questions people ask
How often should FTP be retested?
Every six to ten weeks is the usual rhythm, and more often than that mostly measures how fresh you are rather than how fit. A maximal test is itself a hard session that costs a day or two of recovery, so it needs to earn its place in the week. Many riders now skip formal tests entirely and infer the number from real efforts: a hard 40 to 60 minute climb or a race gives data that is at least as good, at no extra cost. What matters more than the retest interval is consistency of method, because a jump in FTP that came from switching protocols is not fitness.
My ramp test gives a much higher FTP than my 20-minute test. Which one is right?
Probably the lower one, if your threshold sessions built from the higher number keep failing. The ramp protocol ends in a short maximal effort, so riders with a strong anaerobic contribution finish at a high one-minute power that does not reflect what they can hold for an hour. Sprinters and track riders often see this gap; diesel-engined endurance riders sometimes see the reverse. The practical test is behavioural rather than numerical: set the zones from a number, ride a 2 x 20 at 95 to 100 percent of it, and if you cannot complete the second interval at the prescribed power, the number is too high.
Is sweet spot a real zone?
It is a label for the overlap between the top of tempo and the bottom of threshold, roughly 88 to 94 percent of FTP, and it exists because that band gives a large share of the adaptation of threshold work at a noticeably lower recovery cost. It is not a separate physiological state. Its popularity with time-limited riders is straightforward: it is the highest intensity that can be repeated several times a week without the recovery debt accumulating. Its risk is the same thing viewed differently, because it is comfortable enough to do constantly and doing it constantly crowds out both the easy volume and the genuinely hard work.
Why does my power drop on the same route in summer?
Heat is the usual answer, and it is a real physiological cost rather than a lack of effort. Blood is diverted to the skin for cooling, cardiac output serving the muscles falls, and sustainable power falls with it, sometimes by five to ten percent or more in serious heat before any dehydration is involved. Altitude does the same thing through a different route. Neither is captured by a zone table, which is why the zones describe what to aim at and the session tells you what was actually available. Adjusting targets downward on a hot day is normal practice, not a concession.
Do I need a power meter to use this?
For the zones to mean anything as watts, yes, since they are defined as fractions of a measured power figure. A smart trainer with an accuracy specification does the same job indoors. Without either, the equivalent structure exists on the heart rate side and the same caveats apply more strongly there, because heart rate drifts. What you should not do is take an FTP figure from an online estimate based on weight and self-rated fitness and then train to it precisely; an estimated FTP carries an error band far wider than the width of the zones you would build from it.