Functional threshold power is the number most structured cycling training is built around, and it is defined in watts. That creates an obvious problem for the large share of age-group cyclists and triathletes who train without a power meter: the metric everyone talks about seems to require hardware they do not own.
It does not. FTP describes a physiological ceiling, and watts are only one way to read it. Heart rate, speed on a known climb, trainer resistance curves and perceived effort are others. This guide covers four field methods, how to run each one, and where each one fails.
Can you measure FTP without a power meter?
Yes, though not directly. FTP is a power value, but the threshold it describes can be found without one. The most reliable options are a 30-minute heart rate test for lactate threshold heart rate, a timed climb converted to watts with basic physics, virtual power on a trainer, and calibrated perceived exertion.
The distinction matters. To train at the right intensity, you need a reliable anchor for threshold, not a watt number, and heart rate or perceived exertion can provide that. If you want watts specifically, the climb method is the only one of the four that produces an outdoor estimate you can reasonably defend.
What does FTP actually measure?
FTP was popularised as a practical field proxy for the power output at your lactate threshold, specifically the second threshold, sometimes called maximal lactate steady state. Below it, lactate production and clearance stay roughly balanced and you can hold the effort for a long time. Above it, metabolic byproducts accumulate, breathing becomes laboured, and time to exhaustion shrinks quickly. For most trained riders, FTP corresponds to an effort sustainable for somewhere between 40 and 70 minutes.
This threshold exists regardless of what you measure it with. Your heart, your breathing and your speed up a hill all respond to crossing it. A power meter simply reports the mechanical output with high precision and zero lag, which is why it became the default. The methods below are different instruments pointed at the same turning point; what changes between them is precision, lag and susceptibility to heat, wind and fatigue.
How do you find your threshold heart rate?
Lactate threshold heart rate (LTHR) is the most widely used power-free anchor, and the protocol popularised by coach Joe Friel remains the standard field test.
The 30-minute LTHR test:
- Warm up for 15 to 20 minutes, including two or three short efforts of around one minute at a hard but controlled intensity.
- Ride 30 minutes solo at the hardest effort you can sustain evenly for the full duration. Treat it like a time trial: no group rides, no drafting, no surges.
- Press the lap button at the 10-minute mark.
- Your LTHR is your average heart rate across the final 20 minutes.
The first 10 minutes are excluded because heart rate lags behind effort. It takes several minutes for it to rise and settle at a level that reflects the work you are doing, so including the opening section would pull the average down.
Run the test on a steady road or a long, consistent climb, or indoors on any trainer. The speed and resistance do not matter here; only heart rate does. Use a chest strap rather than a wrist-based optical sensor, which can lose accuracy at high intensities and with the wrist movement of hard riding.
Once you have LTHR, Friel's cycling zones translate it into training intensities:
| Zone | % of LTHR | Purpose |
|---|---|---|
| Zone 1 | Below 81% | Recovery |
| Zone 2 | 81 to 89% | Aerobic endurance |
| Zone 3 | 90 to 93% | Tempo |
| Zone 4 | 94 to 99% | Sub-threshold |
| Zone 5a | 100 to 102% | Threshold |
| Zone 5b | 103 to 106% | Aerobic capacity |
| Zone 5c | Above 106% | Anaerobic capacity |
For an athlete with an LTHR of 168 bpm, Zone 2 sits between roughly 136 and 150 bpm, and threshold work sits between 168 and 171 bpm.
Where heart rate falls short. Heart rate is a strong anchor for steady efforts and a weak one for short intervals. On a three-minute VO2 max repeat, heart rate may still be climbing when the interval ends, so it tells you very little about whether you hit the target. For anything shorter than about five minutes, pace those efforts by feel and use heart rate only to confirm you are recovering between them.
Heart rate also drifts. On long efforts, especially in heat or when underhydrated, it creeps upward at the same output as blood volume shifts and core temperature rises. Fatigue can push it either way: a deeply tired athlete often cannot reach normal threshold heart rate at all, which is a useful warning sign in itself. Caffeine, sleep and illness all add noise.
One more caution for triathletes: LTHR is sport-specific. Running threshold heart rate is commonly several beats higher than cycling, because running engages more muscle mass. Test each discipline separately rather than reusing one number.
Can you calculate FTP in watts from a climb?
Yes, and it is the most underused method on this list. On a steady climb, the power you produce goes almost entirely into lifting your body and bike against gravity. Gravity is easy to calculate precisely, which means a timed climb can be converted into a credible watt estimate with nothing more than a scale, a reliable gradient figure and a stopwatch.
Total power on a climb is the sum of three demands, divided by drivetrain efficiency:
- Gravity: total mass × 9.81 × gradient × speed
- Rolling resistance: a small, roughly fixed fraction of weight
- Aerodynamic drag: grows with the cube of speed, so it is minor at climbing speeds
Worked example. A rider weighing 75 kg, with bike, kit, shoes and bottles adding 9 kg, rides a 5 km climb averaging 7 percent in exactly 20 minutes. That is 15 km/h, or 4.17 metres per second. Using standard assumptions (rolling resistance coefficient of 0.005, a seated climbing drag area of 0.40 m², air density of 1.2 kg/m³ and 97.5 percent drivetrain efficiency):
- Gravity: about 246 W
- Rolling resistance: about 18 W
- Aerodynamic drag: about 18 W
- Total: about 281 W for 20 minutes
Gravity accounts for roughly 87 percent of the total. That is exactly why the method works: the term you can calculate most precisely dominates, and the terms that require guesswork are small. Applying the standard 95 percent conversion used for a 20-minute test gives an estimated FTP of about 267 W.
Choosing the right climb. The method only holds up under specific conditions:
- Gradient of 6 percent or steeper. On shallower roads, speed rises and aerodynamic drag becomes a large, uncertain share of the total.
- Consistent gradient. Avoid climbs with flat sections or descents, which distort average speed.
- 15 to 25 minutes long. Long enough to approximate threshold, short enough to pace evenly.
- Still air. Wind is the variable you cannot correct for.
- No drafting. Climb alone
Where the climb method breaks. The inputs that matter most are not the ones most riders worry about. Using the worked example above:
- A gradient figure that is off by just 0.5 percentage points shifts the estimate by about 18 W, roughly 6 percent.
- A total mass that is off by 2 kg shifts it by only about 7 W.
- An unnoticed 10 km/h headwind adds roughly 32 W of real demand that the calculation never sees.
The lesson: weigh yourself and your bike carefully, but obsess over the gradient. Signposted gradients are often rounded, and segment data can be noisy. Where possible, calculate the average yourself from elevation gain divided by distance (350 m of gain over 5 km is 7 percent), using a source with good elevation data. Then repeat the test on the same climb in calm conditions. Even if the absolute number carries a few percent of error, the same climb ridden under the same conditions tracks change over time very reliably.
How accurate is virtual power on a basic trainer?
If you ride a wheel-on trainer without built-in power measurement, apps can estimate power from wheel speed using the trainer's known resistance curve. This is usually called virtual power, and it lets you run a standard ramp or 20-minute FTP test indoors.
The method is only as good as the match between your setup and the curve the software assumes. Tyre pressure, roller tension, tyre temperature and unit-to-unit variation all change the real resistance at a given speed, so the absolute number can be off by a meaningful margin in either direction.
Its real value is consistency. To make virtual power useful:
- Use a dedicated trainer tyre at the same pressure every session.
- Set roller tension identically each time (count the turns on the adjustment knob).
- Warm up for 10 minutes before any test, since resistance on fluid and magnetic units changes as they heat.
- Treat the result as an indoor-only number for that specific setup.
Under those conditions, virtual power tracks your progress and sets indoor training zones well. What it should not do is set expectations for outdoor riding or for a future power meter, which may read quite differently.
It is also worth checking what you already own. Many direct-drive smart trainers measure power directly, and if yours does, you have a power meter indoors and can run any standard FTP test protocol without estimation.
Is perceived exertion good enough for threshold training?
Rating of perceived exertion (RPE) is the oldest intensity tool in endurance sport, and it is more robust than its reputation suggests. Your perception of effort integrates breathing, muscular sensation and core temperature into a single signal, and it responds instantly, without the lag that limits heart rate.
On a 10-point scale, threshold sits at around 7: hard, sustainable but uncomfortable, with breathing deep and rhythmic. The talk test makes it more concrete:
- Below the first threshold (true Zone 2 territory), you can speak in full sentences.
- Between the thresholds, speech drops to a sentence at a time.
- At threshold, you can manage a few words between breaths.
- Above threshold, speech stops being worth the effort.
RPE is most useful for short, hard intervals where heart rate cannot keep up, and for keeping easy days genuinely easy, which is where most age-group athletes go wrong. In a polarized week, the bulk of riding should feel conversational, and RPE is often a better guard against creeping intensity than any device.
Its weakness is subjectivity. Perception shifts with motivation, heat, sleep and caffeine, and inexperienced athletes tend to underestimate early in an effort. RPE becomes reliable through calibration: rate every session, compare it against heart rate on steady efforts, and your sense of threshold sharpens within a few weeks.
Which method should you use?
| Method | What you get | Equipment | Main weakness | Best for |
|---|---|---|---|---|
| 30-minute LTHR test | Threshold heart rate and zones | Chest strap heart rate monitor | Lag and drift; useless for short intervals | Steady endurance and tempo training |
| Climb calculation | Outdoor watt estimate | Scale, timing device, reliable gradient data | Gradient and wind errors | A defensible watt figure and outdoor progress tracking |
| Virtual power | Indoor watt estimate | Wheel-on trainer, speed sensor, app | Setup-dependent absolute accuracy | Structured indoor training |
| RPE and talk test | Real-time threshold sense | None | Subjective, needs calibration | Short intervals and easy-day discipline |
The strongest approach is triangulation. Run the LTHR test to anchor your zones. Ride a calibrated climb every four to six weeks to put a watt estimate on your progress. Use RPE for anything shorter than five minutes and as a cross-check on everything else. When two methods agree, you can trust the number. When they diverge, that divergence is information: a threshold heart rate you cannot reach, combined with a slower climb, points to fatigue rather than lost fitness.
If your rides sync to Triforge from Garmin, Wahoo, Suunto or Strava, Training Analytics makes this kind of cross-checking easier, because a falling heart rate at the same climb time or the same perceived effort is often the earliest visible sign that your threshold has moved.
What changes when you get a power meter?
Expect your first power-based FTP test to disagree with your estimates, and do not read too much into the gap. A climb estimate within 5 to 8 percent of a power meter reading is a good result, and virtual power figures may be further out in either direction.
What carries over is your understanding of threshold. Athletes who have trained by heart rate and feel tend to pace tests better and notice fatigue earlier. Power adds precision; it does not replace the other signals.
Frequently asked questions
Is a heart rate FTP test accurate?
A 30-minute test gives a reliable lactate threshold heart rate for setting training zones, provided it is paced evenly and run with a chest strap. It does not produce a watt value, and heat, fatigue and caffeine can shift the result by several beats, so test in consistent conditions.
Can Strava estimate my FTP without a power meter?
Strava shows estimated power on rides without a power meter, calculated from speed, elevation and your entered weight. Because it cannot account for wind, drafting or road surface, those figures are rough. A dedicated solo test on a steady climb of 6 percent or steeper is far more reliable.
How do I convert threshold heart rate into watts?
There is no direct conversion, because the relationship between heart rate and power differs between athletes and shifts with fitness, heat and fatigue. To get a watt estimate without a power meter, use a timed climb calculation or virtual power on a trainer, and keep heart rate as your zone anchor.
Does a smart trainer count as a power meter?
Many direct-drive smart trainers measure power directly and can be used for any standard FTP test. Basic wheel-on trainers without power measurement rely on virtual power, which estimates watts from wheel speed and a resistance curve, so treat those results as setup-specific.
How often should I retest without a power meter?
Every four to six weeks works for most athletes, ideally at the end of a recovery week so fatigue does not suppress the result. Repeat the same method, on the same climb or trainer setup, in similar conditions, so that changes reflect fitness rather than measurement noise.
The bottom line
A power meter is a precise instrument, not a prerequisite. Threshold heart rate gives you reliable zones, a well-chosen climb gives you a watt estimate you can defend, virtual power keeps indoor training structured, and calibrated perceived exertion handles everything the devices miss. Use them together, retest consistently, and you can train at threshold with far more precision than the absence of a power meter would suggest.
What a 20-minute climb is worth in watts
Worked example: 84 kg rider and bike, 5 km at 7 percent, ridden in 20:00
- 246 WGravity
- 18 WRolling resistance
- 18 WAerodynamic drag
Gravity is 87 percent of the work, and it is the part you can calculate precisely. At 95 percent of the climb average, estimated FTP is about 267 W.
Which input errors move the estimate most
Change in estimated climb power from a single wrong input
Assumes rolling resistance coefficient 0.005, drag area 0.40 m², air density 1.2 kg/m³, drivetrain efficiency 97.5 percent.
