Cycling FTP Calculator: Power Training Zones
FTP (Functional Threshold Power) is the highest power a cyclist can hold in a steady state for roughly an hour. It's the backbone of structured power-based training: every intensity zone is expressed as a percentage of this one number, which replaces heart rate as the primary reference — more responsive, and far less thrown off by heat or how tired you are that day.
Two common ways to estimate it: a maximal 20-minute test, with the average power scaled down to 95% (a 20-minute effort always overstates what you can hold for a full hour), or a direct 60-minute effort, more accurate but a much harder thing to pace solo.
FTP comes from a maximal voluntary effort: it depends on how motivated you are on test day as much as on fitness. Going out too hard early undersells your real FTP; retest under similar conditions (same route, or the same trainer setup) so successive tests are actually comparable.
Your test
Tested FTP versus algorithm-estimated FTP (Zwift-style auto-detection)
FTP is, roughly, the most power you can hold for an hour. This page gets it from a dedicated effort: knock 5% off a full-gas 20-minute average, or take the hour average as-is. What Zwift or TrainerRoad pins on your profile after a few rides is something else. That's a model reading files, not you riding a test.
The thing is, auto-detect looks at peaks, durations, a fatigue curve. It does not know you slept badly, that the garage hit 28°C, or that you sat up with two minutes to go. Against a proper test it often sits 5 to 10% off. Set zones from a high guess and threshold work becomes a war. Set them from a low one and you cruise through Z4 thinking you have more to give.
Well, keep the auto number as a trend line. If it ticks up three weeks running, fitness is probably moving. For the watts you target on Wednesday night, test. Same protocol, same setup.
How often to retest and recalibrate your FTP
A test is a hard session with extra nerves. You do not need one every Sunday.
New to structured training, or coming back from a layoff: every 4 to 6 weeks — the number moves fast. Intermediate: 6 to 8 weeks. Well-trained: 8 to 12 weeks, because genuine change gets slower and a test every fortnight mostly reads freshness.
Retest after illness, a holiday that stopped you riding, a real weight change, or a new meter. Look, testing in the middle of a heavy block tells you how cooked you are. It does not tell you your FTP.
Do not mix protocols when you compare months. A March 20-minute and a June ramp are two different instruments.
The seven Coggan zones in practice: what a typical training week looks like
Coggan's seven bands are just percentages of FTP. Coaches use them so "threshold" means the same thing on two different bikes.
Z1 below 55%: spinning, no load. Z2 from 55 to 75%: endurance you can talk through — most of the week's hours live here. Z3, 75 to 90%: tempo, conversation in pieces. Z4, 90 to 105%: threshold, right around FTP. Z5, 105 to 120%: VO2, a few hard minutes. Z6, 120 to 150%: anaerobic, 30 seconds to three minutes. Z7 above 150%: sprints, skill as much as engine.
A week built around commuting and a Saturday club 10 looks nothing like the "Monday off, Saturday 4 hours" template. Try this instead:
- Monday: commute both ways in Z2 (say 40 minutes each), no heroics at the lights.
- Tuesday: rest, or 25 minutes very easy if the legs feel wooden.
- Wednesday: indoor trainer, 3 × 12 minutes at 88–94% FTP (sweet spot), 5 minutes easy between.
- Thursday: 50-minute spin, Z1–Z2, cadence a bit high.
- Friday: off.
- Saturday: club 10-mile TT, or 6 × 3 minutes up a local drag at 110–120%.
- Sunday: café ride, two hours, keep intensity factor around 0.75 — chat, don't chase every sign.
That split is close to 80% easy, 20% hard. Flip it and you get tired, not fitter.
W/kg versus raw watts: which one matters depends on the terrain
Your meter speaks watts. A climb speaks watts per kilo. On the flat, drag rises roughly with speed squared, so the rider who can put out more raw watts usually holds the speed — body size and all.
Picture a 60 kg climber at 4 W/kg: 240 W. An 80 kg rider at 3.5 W/kg is doing 280 W and still losing time on the hill. Forty extra watts on the head unit, slower to the top. Point the same pair down a dual carriageway into a headwind and the heavier engine often tows.
Keep both numbers. Watts to hit a zone. W/kg to decide whether Box Hill is a workout or a massacre.
Solo FTP versus group-ride FTP: what drafting changes
FTP does not care that someone is on the front. Speed does. Drafting in a bunch often cuts the power for a given speed by 30 to 40%. Three hours at 22 mph sat in: Z2. The same 22 mph on your own: Z3, and Z4 the moment the road kinks up.
Frankly, a lot of "easy" club runs are only easy while the group holds. Once it splits, the wattage you thought was endurance isn't. The other way around too: a twitchy chain gang inflates TSS without building much, because you spent the morning surging rather than riding a line.
Ride alone to learn how to carry the load. Ride in a group to learn how to spend it. Monday's FTP is the same either way. The file is not.
Alternative protocols: the 8-minute test, the ramp test
This page uses 20 minutes × 0.95, or a straight hour. Two shorter protocols show up on indoor trainers and in older plans.
- Warm up thoroughly. Then two all-out 8-minute efforts, 10 minutes easy between them. Average the two, take about 90%. Kinder than a 20-minute — and it can under-read riders who are good at lasting.
- A ramp test steps the watts until you stop. A common estimate is 75% of the last completed step. Short, tidy, easy to repeat on a trainer. It tends to underestimate against a 20-minute, because the fatigue is a different shape.
FTP ≈ 0.95 × P20 min · FTP ≈ P60 min · ramp ≈ 0.75 × last completed step
Pick one and stay with it. Progress is a line, not a mix of gadgets.
Why FTP differs from lactate threshold measured in a lab
A lab threshold is often pinned near 4 mmol/L of blood lactate — the intensity where lactate takes off. FTP is a one-hour job, or the 20-minute stand-in. Different question, different answer.
In most riders FTP sits 5 to 10% below that lab mark. Not a scandal: one is a blood draw, the other is a performance. Lactate also shifts with heat, how long you've been going, and yesterday's ride. Your 250 W from Saturday does not have to match the "4 mmol" step on a printout.
For writing next week's sessions, FTP is enough. Book the lab when you need a diagnosis — unexplained stall, illness, a rider on a talent pathway — not to set three times twelve on the turbo.
Improving FTP: which type of sessions moves the needle most
Harder is not a plan. The sessions that lift FTP are boring to list and effective to ride.
Z2 volume two or three times a week is still the base — capillaries, fat use, economy. Then threshold: two or three blocks of 15–20 minutes at 95–105% FTP. Long tempo (2 × 20 or 3 × 15) suits sportives and anyone who has to stay on for hours. VO2 work, 3 to 5 reps of 3–5 minutes at 110–120%, lifts the ceiling and, later, the threshold.
Honestly, the trap is all intensity, no Z2. Heroic weeks, flat FTP by October. An 80/20 mix (easy / hard) is simply the dose most bodies can absorb without coming apart.
Road, mountain bike and indoor-trainer FTP: where the gaps come from
Same legs, three numbers. Fitness didn't lie. The room did.
Outdoors you often see the highest FTP: moving air, a bit of sway in the bike, something to look at. On an indoor trainer, 5 to 10% lower is common — heat, no bike movement, boredom. Off-road, average power collapses (hike-a-bike, tech, constant cuts) while normalized power can still match a road file.
If you live on the turbo from November to March, set zones there. The first climb in April will feel "easy" at the same watts, and it should. Export a summer road FTP onto the trainer and the first interval will bury you. That's the setup talking, not a lost winter.
Normalized power and intensity factor: going beyond raw FTP
FTP is a ceiling. A ride is a shape. Normalized power weights the surges so the number tracks what the effort actually cost, not the polite average. A lumpy circuit with attacks has an NP well above mean power. Intensity factor is just NP divided by FTP.
IF = NP / FTP
IF 0.75 is endurance that builds. IF 0.95 is a very hard day, threshold-ish for the duration. Comparing a two-hour hilly loop with a 50-minute trainer set without IF is guesswork. Those two figures also feed TSS, which is how you notice that three "short" days were not short.
The 20-minute × 0.95 versus 0.90 debate
Allen and Coggan popularised 95% of a 20-minute all-out as a stand-in for the hour. Plenty of coaches now shave more — 0.93, even 0.90 — especially for riders who are punchy over 20 minutes and fade when the clock goes past 40. A time-triallist who can sit within a few watts of their 20-minute power for an hour may be closer to 0.95. A crit specialist often isn't.
This calculator uses 0.95, because that's the protocol on the form. Changing the factor every other week because a forum thread said so just scrambles your zones.
If 20-minute efforts feel like a different sport from a long climb, run a 40- or 60-minute test when you can, or take 0.90 for a block and reassess. Consistency beats the "right" decimal.
Zero-offset your power meter before you test
Pedals, cranks, spider-based meters: they drift. Temperature, a knock in the car, last week's muddy ride. A zero-offset (or calibration, depending on the brand) takes a few seconds. Skip it and you can be 5–10 W out before you even start the effort. On a 250 W FTP that's a zone error, not a rounding error.
Do it unclipped, no load on the pedals, after the unit has seen the room's temperature for a few minutes. Then leave it. Mid-test "recalibrations" are how people invent a new FTP by accident.
A test with a stale offset is worse than no test. You will write zones to a number that never existed. Zero-offset, ride the protocol, then — and only then — type the average into this page.