Four Ways to Test FTP.

An athlete came to me having done a ramp test in January and a twenty minute test in March. The ramp said 262. The twenty minute said 241.

He wanted to know which one was real, and whether he had lost twenty watts.

Neither, and no. He had used two different rulers and read two different numbers off them, which is exactly what you would expect.

The protocol is part of the measurement. Comparing a number from one against a number from another tells you nothing.

The Twenty Minute Test

Warm up, ride twenty minutes as hard as you can hold evenly, take ninety five percent of the average.

The ninety five percent exists because twenty minutes at maximum is meaningfully above what you could hold for an hour, and the fraction is an attempt to bridge that gap.

What is good about it: it is repeatable, the effort is a realistic simulation of racing, and it produces a number that tends to work when you go and train with it.

What is not: it is genuinely horrible, which means athletes avoid it or quietly sandbag it. And the ninety five percent is an average across a population, not a truth about you. If you are a rider with a big anaerobic contribution, twenty minutes flatters you and your derived FTP will be too high, so your threshold sessions will be undoable and you will conclude you are unfit.

The Ramp Test

Power increases every minute until you cannot continue. FTP is estimated as roughly seventy five percent of the best one minute you managed.

What is good: it takes about twenty five minutes total, it is far less unpleasant, and athletes will actually do it regularly, which is worth more than people credit.

What is not: it is heavily influenced by anaerobic capacity, which has almost nothing to do with the hour you are trying to estimate. Punchy riders with a big top end get a flattering number. Diesel riders with a strong aerobic engine and no sprint get an unflattering one, and they are usually the athletes for whom threshold matters most.

That is where the twenty watt gap came from. He is punchy, so the ramp rewarded him.

The Eight Minute Test

Two eight minute efforts with ten minutes between, take ninety percent of the average of the two.

It exists mostly because twenty minutes is hard to pace and eight is not. The trouble is that eight minutes is well within the range where anaerobic contribution matters, so it carries the same distortion as the ramp with more suffering than either.

I do not use it. If you are going to accept an anaerobic distortion, take the ramp and save yourself the misery.

The Full Hour

Ride as hard as you can for sixty minutes. That number is your FTP by definition, with no correction factor and no assumption.

It is the most accurate and almost nobody should do it. It is brutal, it requires a course or a turbo you can hold for an hour without interruption, it takes days to recover from, and it is so unpleasant that athletes will not repeat it every eight weeks, which is what makes a test useful.

A perfectly accurate number you take once a year is worth less than a slightly imperfect one you take six times.

What I Actually Use and Why

The twenty minute test, for most athletes, most of the time.

Not because it is the most accurate. Because it hits the balance I care about: close enough to be useful, repeatable enough to track, and similar enough to a race effort that it also teaches pacing.

The exception is athletes new to structured training, who cannot yet pace twenty minutes without blowing up in the first five. For them I use a ramp for the first couple of blocks, then move them across once they can pace.

And the rule that matters more than the choice: whichever you pick, never change it. All you actually care about is the trend, and a trend requires the same ruler each time. Switching protocols mid-season destroys your ability to know whether you improved.

Making the Number Mean Something

Test under conditions you can repeat. Same time of day, same bike, same trainer, same fan, same warm up, same fuelling. Cooling matters more than people think indoors, and a hot room can cost you a genuine ten watts.

Do not test on the back of a heavy week or a bad night. You will measure your fatigue, set every zone in the following block too low, and train soft for eight weeks.

And sense-check the result against your training. If your new FTP makes your threshold intervals impossible, it is too high. If sweet spot feels easy, it is too low. The sessions are a better validator than the test, and I will adjust a number by a few percent based on how the first week of a block actually goes.

The Bottom Line

Pick one protocol, ideally the twenty minute, and never change it. Control the conditions. Test when fresh, every six to eight weeks.

And stop comparing your number to anybody else's, including your own from a different protocol. It is a ruler, and only the direction it moves in matters.