How often should you test your FTP?

Testing FTP feels productive. It produces a number, the number goes up or down, and it seems like information. Most of the time it is not — it is a hard session spent confirming something your ordinary training was already telling you.

A sensible default cadence

There is no established interval at which threshold power needs remeasuring — the figures below are coaching guidelines, not a physiological schedule. What they reflect is how fast you are likely to be changing, which depends mostly on how long you have been training:

If you are setting a number for the first time, the FTP calculator estimates it from a 20-minute test, and ramp test vs 20-minute test covers which protocol to use. Whichever you pick, keep using the same one — comparing a ramp result against a 20-minute result tells you about the protocols, not about your fitness.

Why most riders test too often

Three reasons the monthly test habit tends to cost more than it returns.

A test is a hard session that produces no training benefit you could not get better elsewhere. A maximal effort plus the recovery it demands is roughly a quality day, and a threshold workout would have developed more fitness for the same cost.

The measurement is noisier than the change. Sleep, caffeine, room temperature, motivation, and how well you pace the effort all move a test result by several watts in either direction. Over eight weeks of good training a trained rider might genuinely gain five to ten watts — the same order as the noise. Test monthly and you are mostly measuring your mood.

The number is a zone-setting tool, not a score. Its only job is to put your training zones in roughly the right place. A five-watt error moves a threshold interval target by about five watts, which changes essentially nothing about the session. Chasing precision here is effort spent on the wrong problem.

Signals that your FTP setting is actually wrong

Better than a calendar: retest when your training starts contradicting your number. The signals are reasonably clear in both directions, as long as you look for a pattern rather than a single session.

Set too low — threshold intervals feel comfortably repeatable, you finish the final rep feeling you could do two more, you drift above the target without intending to, and sweet-spot work feels like endurance riding. If that describes three or four consecutive quality sessions, your zones have fallen behind your fitness.

Set too high — you cannot complete sessions as written, the last intervals fall well short of target, and work that should be sustainable feels maximal. The complication here is that this also describes accumulated fatigue, so before concluding your FTP is wrong, check whether you are simply under-recovered. Take a genuinely easy week first; if quality sessions still fall apart afterwards, the setting is the problem.

Something structural changed — a long break, illness, a significant weight change, or moving to a different bike or power meter. Each is a legitimate reason to reset rather than adjust.

What not to read as an FTP change

Plenty of things make a workout feel harder or easier without your threshold having moved at all:

• Heat. Riding indoors in a warm room, or outdoors on the first hot day of the year, raises cardiovascular strain at the same power. That is a cooling problem, not a fitness one.
• Sleep debt. A run of short nights reliably degrades session quality. One night rarely does.
• Fueling. Starting a threshold session under-carbohydrated makes the last intervals disproportionately hard.
• Where you are in a block. Sessions in the third week of a build are supposed to feel harder than the first. That is the block working.
• Life stress. Fatigue from outside cycling still arrives in your legs, and it does not distinguish itself from training fatigue.

The practical filter: if a bad session has an obvious explanation, keep your number and move on. If several bad sessions have no explanation, act — though the first action is usually a recovery week, not a test.

Alternatives to a formal test

For most riders training regularly, formal testing can be replaced almost entirely by paying attention:

• Use your hardest completed session. If you finished 4x10 minutes at your threshold target and it was genuinely hard but controlled, your setting is about right. That is the same evidence a test provides, gathered without spending a day on it.
• Adjust in small steps. If sessions have been consistently too easy for several weeks, raise your FTP by two or three percent and see whether the next block still lands. This tracks fitness more smoothly than stepwise jumps every time you test.
• Use a long, hard group ride or event. A well-paced climb or a hard 40 to 60 minutes in a race often produces a better estimate than a solo test, because motivation is rarely the limiter.
• Let modelled estimates do the routine work. Most platforms will infer a threshold from your recent hard efforts. It is imprecise, but so is testing, and it costs nothing.

A formal test still earns its place occasionally: when you have no number at all, when you are coming back after a long layoff, or before a block where the prescription depends on getting it right. The rest of the time, your training is the test.

How SmarterTraining thinks about this

SmarterTraining builds each session around your current FTP, so keeping it roughly current matters — but roughly is the operative word. Rather than holding you to a testing schedule, it adjusts session difficulty from how your recent training has actually gone, which means a slightly stale number degrades gracefully instead of quietly making every workout wrong.

Takeaway

Takeaway: If you want a default, roughly every 8 to 12 weeks in stable training — but it is a rule of thumb, not a physiological interval. Retest on evidence rather than the calendar: a run of sessions that are clearly too easy or clearly impossible. And remember the number only has to be close enough to put your zones in the right place.