FTP for Cyclists: The 20-Minute Test, Power Zones, and W/kg Explained
Estimate cycling FTP from a 20-minute test (×0.95), build the seven Coggan power zones in watts, and read your watts per kilogram with a worked example.
FTP for Cyclists: The 20-Minute Test, Power Zones, and W/kg Explained
If you train with a power meter, one number sits underneath everything you do: your Functional Threshold Power, or FTP. It is the highest average power, in watts, that you can hold for roughly an hour. Almost every structured workout is written as a percentage of it. Without a current FTP, "ride at threshold" or "do four sets at tempo" are just words. With one, they become specific watt targets you can pin a session to.
The catch is that a full-hour all-out test is genuinely miserable, and almost nobody paces it well. So instead of riding for sixty minutes at your limit, the cycling world settled on a shorter proxy: the 20-minute test. This post walks through how that test turns into an FTP, how FTP fans out into the seven Coggan training zones, and what watts per kilogram tells you that raw watts cannot.
Why the 20-minute test works
Here is the one fact that the whole method rests on: your FTP is about 95% of your best 20-minute average power. You can push noticeably harder over 20 minutes than you can over a full hour, so the 20-minute figure overshoots true threshold by roughly 5%. Multiply it by 0.95 and you land back at a number you could realistically defend for an hour.
The protocol matters as much as the math. Warm up properly, including a short hard effort to open the legs. Then ride a maximal, evenly paced 20 minutes and record the average power. The word that does the heavy lifting is evenly. Treat it like a time trial, not a sprint that detonates at minute five. A test that starts at 350 W and limps home at 220 W produces a flattering average and an FTP you can never actually hold, which then poisons every zone built on top of it.
You can run this on a climb, a flat road in still air, or a smart trainer. The trainer is easiest to control because there is no traffic and no wind to break your rhythm.
A worked example, start to finish
Say you finish the test and your head unit reads a 20-minute average of 260 W. Apply the formula:
FTP = 260 × 0.95 = 247 W
That 247 W is now the anchor for your zones. Drop the same 260 W into the cycling FTP calculator and it returns the FTP and the full zone table instantly, but it is worth seeing where the bands come from. Here are the seven Coggan zones for an FTP of 247 W, each one a percentage range of that anchor:
| Zone | Name | % of FTP | Watts | |------|------|----------|-------| | Z1 | Active recovery | under 55% | under 136 W | | Z2 | Endurance | 55–75% | 136–185 W | | Z3 | Tempo | 75–90% | 185–222 W | | Z4 | Lactate threshold | 90–105% | 222–259 W | | Z5 | VO2max | 105–120% | 259–296 W | | Z6 | Anaerobic capacity | 120–150% | 296–371 W | | Z7 | Neuromuscular sprint | over 150% | over 371 W |
Notice the ranges are contiguous: where one zone ends, the next begins, so there are no gaps to fall into. The headline takeaway is that every training zone is just a slice of your FTP. Change the FTP and the whole table redraws. This is exactly why a stale FTP is so quietly damaging. If you tested at 247 W three months ago and you are fitter now, your "threshold" intervals are sitting too low and improving nothing.
What each zone is for
The bands are not arbitrary; each one trains a different physiological system.
- Z1 / Z2 (under 75%) build your aerobic base. Z2 in particular is where most of your weekly hours should live. At 136–185 W in our example you can hold a full conversation, fat oxidation is high, and you grow mitochondria and capillaries without deep fatigue.
- Z3 (tempo) is the "comfortably hard" middle. Useful in moderation, but the trap is riding everything at a grey medium-hard pace that is too easy to drive adaptation and too hard to recover from.
- Z4 (lactate threshold) is your FTP work itself. A session like "4 × 8 minutes at threshold" lives here, around 222–259 W.
- Z5 (VO2max) is the short, lung-searing stuff that lifts your ceiling. At 259–296 W these are minutes, not tens of minutes.
- Z6 / Z7 are anaerobic and sprint efforts, measured in seconds.
A polarized approach keeps roughly 80% of your time in Z1–Z2 and reserves the hard 20% for genuine Z4–Z5 intervals. Power zones pair naturally with your heart-rate bands, and you can sanity-check the two against each other with the heart rate zones calculator so a session lined up by watts isn't contradicted by a wildly different effort signal.
Watts per kilogram: the number that matters on a climb
Raw watts win on the flat. The moment the road tilts up, what matters is watts per kilogram — your FTP divided by your body weight. Gravity does not care how many total watts you make; it cares how many you make per kilo you have to haul uphill.
Take two riders. One holds 300 W at 85 kg. The other holds 250 W at 65 kg. By raw power the first rider looks stronger. But the W/kg tells a different story: 300 ÷ 85 = 3.53 W/kg versus 250 ÷ 65 = 3.85 W/kg. On a long climb the lighter rider pulls away despite making fewer watts.
Rough male brackets help you place yourself: under 2 W/kg is untrained, 2–3 recreational, 3–4 a solid club rider, 4–5 a strong amateur racer, and above 5 W/kg approaches elite. The same labels sit slightly lower for women. Our 247 W rider at, say, 70 kg comes out at 3.5 W/kg — a respectable club-level figure.
How I actually use this
I retest every four to six weeks during a focused block, and the first time I did it properly the value of the 0.95 step finally clicked. My 20-minute average came in at 268 W and I was ready to set my threshold intervals at 268. Taking 95% dropped that to 255 W, and the difference was night and day on the trainer — at 268 W targets I was blowing up on the third interval, while at 255 W I could finish a clean set and actually repeat it the next week. The haircut is not pessimism; it is the number you can defend. Now I treat a fresh test as a maintenance task, drop the new average in, let every zone redraw around it, and trust that the targets are honest rather than aspirational.
A few mistakes to avoid
- Sprinting the test instead of pacing it. Even effort start to finish, or the average lies and the whole chart sits too high.
- Treating the 20-minute average as your FTP. It is not. FTP is 95% of it. Skip the 0.95 and every zone runs about 5% hot, and your easy rides creep into tempo.
- Forgetting to retest. Zones built on an old FTP make a fitter rider train too easy. If threshold intervals start feeling comfortable, the number is stale.
Run your own test, take 95% of the average, and let the cycling FTP calculator lay out the zones and your W/kg in one pass. Then go pin a workout to a real number instead of a feeling.
Made by Toolora · Updated 2026-06-13