You finish a group ride feeling completely cooked — legs gone, lungs burning — while the guy next to you is chatting like he just woke up from a nap. Same route, same hills, same wind. Different result. You wonder if you trained harder this week or just trained wrong.
That gap usually comes down to one thing: you are riding on guesswork, and he is riding on data. Not fancy data. Just one number that tells you exactly how hard you can push before your body starts falling apart — your Functional Threshold Power, or FTP.
Maybe you ignored FTP testing for years because it sounded like something only racers cared about. And now your riding paid the price. In this post, we will cover what FTP actually is, why it matters even if you never pin on a race number, who genuinely benefits from testing it, and how to do it without dreading the whole process.
Table of Contents
What is FTP?
You finish a ride feeling like you left everything on the road — then a faster rider cruises past like they’re on a Sunday stroll. Same route, same conditions. What gives?
Chances are, they know something about their body that you don’t yet: their Functional Threshold Power (FTP). FTP is the highest average power output a cyclist can sustain for 60 minutes. It’s not your sprint peak — it’s your engine’s sustainable ceiling, measured in watts.
FTP measures the highest average power you can sustain for roughly 60 minutes — not your sprint peak, but your engine’s sustainable output. It gives you a personal number, in watts, that anchors every workout to your actual fitness level rather than a generic plan copy-pasted from the internet. It’s a personal number. Two riders can produce identical speed on flat ground while one is working at 65% of FTP and the other at 95%. That’s the gap your post describes.
One nuance worth adding: FTP is closely related to lactate threshold, but they’re not identical. FTP is highly correlated with endurance performance — the higher it is, the harder you can push for longer — but it’s not a direct lab measurement of lactate threshold. It’s a field-practical proxy.
Why do I need FTP Testing?
There are many reason for a FTP test, if one is interested in finding out:
- Track real progress: Without a baseline, you can’t tell if training is working. Retest every 6–8 weeks and the numbers tell you exactly where you stand.
- Pace smarter: Ever blown up halfway through a long ride? Your FTP sets realistic power targets that prevent those painful bonks — and keeps easy days actually easy.
- Simplify your training zones: Once you have your FTP, calculating personalized zones takes minutes. No coach required, no engineering degree needed.
This is a key indicator to train for maximum efficiency and power output.
How do I do FTP Testing?
To get a (relatively) estimate, do the following :
- Use the 20-minute test: Ride all-out for 20 minutes, then multiply your average power by 0.95. It’s not comfortable, but it’s reliable and repeatable.
- Warm up properly: Give yourself 15–20 minutes of easy spinning with a few short hard efforts. A lazy warm-up tanks your result.
- Control your conditions: Use the same route or trainer each time — wind, elevation, and fatigue all skew the number.
Your action for today: open your cycling app and schedule one FTP test this week — treat it like an appointment you can’t cancel.
The Four Test Protocols in 2026
There is no single “best” FTP test protocol. The right one depends on your rider profile and what you’re willing to suffer through.
- Coggan 20-Minute Test (what your post describes) Ride all-out for 20 minutes, multiply average power by 0.95. The 0.95 corrects for the fact that 20-minute power runs slightly above true 60-minute threshold. Works well for most cyclists. Consistency matters more than absolute accuracy. RunBikeCalc
- Ramp Test (the low-friction option) Incrementally increases resistance in one-minute stages until you can no longer continue. It’s a test to failure, well-suited to smart trainers with erg mode. Available on Zwift, TrainerRoad, and Wahoo SYSTM. On Zwift specifically, it starts at 100 watts and increases by 20 watts each minute. You’re not supposed to finish it — the app extrapolates your FTP from your best one-minute effort using proprietary algorithms.
- 8-Minute Test Best suited for riders with higher anaerobic contribution. Two 8-minute all-out efforts, average the power of both, multiply by 0.9. More accessible than the 20-minute version but requires precise pacing judgment.
- Friel 30-Minute Test Less common but used by some coaches. Ride 30 minutes all-out; the average power is your FTP (no multiplier needed). Four protocols every serious rider should know — the Coggan 20-minute, the Friel 30-minute, the ramp test, and the 8-minute.
Trade-off: the ramp test favors riders with high-end anaerobic capacity, so pure diesel engines (strong threshold, weak sprint) will often get a slightly underestimated result.
What's New in 2026: AI-Detected FTP
Three platforms lead the field.
- Garmin Edge auto-detect FTP — available on the Edge 540, 840, and 1050 since firmware v18.0 — scans your rides for sustained efforts of 18–60 minutes and updates an estimated FTP if a new high is detected. It’s reasonable for tracking fitness trends but less reliable for setting precise zones.
- The intervals.icu Critical Power model fits a power-duration curve to your last 90 days of data and outputs both Critical Power and estimated FTP — it tends to read 2–4% higher than a controlled 20-minute test because it uses your best efforts, which often come from racing.
- WKO5’s modelled FTP is used by professional coaches and is most accurate with at least three months of varied power data.
The key limitation: AI-detected FTP drifts with race effort and route topology in a way a controlled test never does. A field test you can replicate every six to eight weeks gives a clean signal across a season. Use auto-detect for passive monitoring between scheduled tests, not as a replacement.
How Often to Test
During base training, test every 12 weeks. More frequent testing causes unnecessary fatigue without meaningful changes. Always test when returning from a break or starting a new training plan. Most coaches land on 6–8 weeks during build phases — your post’s recommendation is solid.
The Gear Question
You want to test your FTP. How do you get some reliable data?
You need a calibrated power meter — pedal-based units, crank-based meters, or a direct-drive trainer. Zero-offset before every test. Aim for a unit with ±1% accuracy or tighter, paired with a head unit or app that records 1-second power.
No power meter? Heart rate can approximate zones, but it’s significantly less precise — lag, heat, caffeine, and fatigue all skew it. If budget is a barrier, a smart trainer is the most cost-effective single purchase for indoor FTP testing.
The Warm-Up
A bad warm-up is the most common reason a test result doesn’t match how you actually ride. Standard protocol is 15–20 minutes easy spinning, then 2–3 short openers at 100–110% of estimated FTP (10–15 seconds each), then 5 minutes easy before the effort begins.
Zones: What to Do With the Number
The Coggan 7-zone model is the most widely used in 2026, supported across TrainingPeaks, intervals.icu, Zwift, TrainerRoad, Wahoo SYSTM, and the Garmin Connect ecosystem.
Once you have FTP, every app calculates zones automatically. Zone 2 (endurance) typically runs 56–75% of FTP. Threshold work sits at 91–105%. Most riders spend too much time in the middle — too hard to recover, too easy to drive adaptation.
Indoor vs. Outdoor FTP Discrepancy
Most riders produce less power on a trainer than on the road — typically 5–8% less, sometimes up to 10% if you run hot or hate the boredom – check out this Ask The Pedalist podcast episode on indoor vs outdoor training to better understand your training options. The reasons are real: no airflow means your core temperature climbs faster, you can’t shift body weight the way you do on a real bike, and the mental cues that push you harder outdoors simply aren’t there.
This means two things. First, don’t panic if your trainer FTP looks lower than what you feel capable of on the road — it probably is, and that’s normal. Second, pick one environment and stick to it. Test indoors, train indoors with those zones. Test outdoors, use outdoor data for outdoor rides. Mixing them scrambles your zones and leads to exactly the guesswork this whole piece is trying to fix.
W/kg: The Number That Puts Watts in Context
200 watts sounds impressive until you’re riding next to a 60kg climber doing the same output. Raw watts measure engine size. Watts per kilogram measures what that engine can do relative to the weight it’s moving — and on any climb, W/kg is the number that decides who rides away.
The math is simple: divide your FTP by your body weight in kilograms.
| W/kg | Rider Level |
|---|---|
| 1.5–2.5 | New or recreational rider |
| 2.5–3.5 | Regular club rider |
| 3.5–4.5 | Competitive amateur |
| 4.5–5.5 | Cat 1 / elite amateur |
| 5.5+ | Professional / WorldTour |
A 75kg rider with a 250W FTP sits at 3.3 W/kg — solid club level. That same 250W on a 65kg rider is 3.8 W/kg — a different tier entirely. Knowing your W/kg tells you far more about your climbing potential and overall fitness ceiling than the watt number alone.
Auto-Detected FTP Is Not a Tested FTP
If you use a Garmin Edge, Strava, or intervals.icu, you’ve probably seen an FTP estimate update automatically after a hard ride. It feels like the work is done for you. It isn’t.
Auto-detection scans your recent rides for sustained high-effort segments and estimates where your threshold sits. It’s useful for passive fitness tracking between tests — but it has real limitations. It can only go up (Garmin’s algorithm won’t flag a decline in fitness), it skews higher when your hardest efforts come from races or group rides where adrenaline and drafting inflate your numbers, and it has no control over conditions, fatigue, or whether you actually buried yourself or just had a good day.
Use auto-detect as a directional signal. Use a proper scheduled test as your actual number. If your app’s auto-estimate and your last test are more than 5–7% apart, that’s your cue to retest — not to update your zones based on an algorithm’s guess.
Nutrition and Recovery Before the Test
FTP testing is a max-effort protocol. Show up depleted or wrecked and you’re not testing your fitness — you’re testing your ability to suffer on empty.
48 hours before: No hard training. Easy spinning or full rest only. Fatigue from your last interval session will clip your result by more than any warm-up routine can recover.
The day before: Normal eating. Don’t carb-load like it’s a race — just don’t restrict either. Stay hydrated.
Test day: Eat a normal pre-ride meal 2–3 hours before. Carbs, moderate protein, low fat and fiber. The same thing you’d eat before a hard training ride. Avoid testing fasted — your 20-minute power will suffer, and you’ll set a false baseline that sandbaggs every zone you train in for the next 6–8 weeks.
Caffeine: Fine. Use what you normally use. Introducing it for the first time on test day is a variable you don’t need.
The goal is to walk into the test with one variable: your fitness. Everything else should be controlled and repeatable.
Your FTP Number Is Waiting — Go Find It
FTP testing is not about chasing a number to brag about. It’s about finally riding with a map instead of guessing in the dark. Once you know your threshold, every ride has a purpose — easy days feel easier, hard days feel honest, and your fitness actually builds instead of stalling. I spent two years riding hard and going nowhere before I ran my first test. The difference was immediate.
Pick one test format — the Ramp Test if you want low friction, the 20-minute if you want precision — and schedule it for this week. Not someday. This week.
