top of page

Stop Doing FTP Tests!

We all know of a rider who has a huge FTP of 350, 375, or even 400+. But we also know of a rider who, despite their incredibly high FTP, fails to perform at even the local level races against riders much weaker than them. This could be a result of everyone's infatuation with finding a single number to quantify their fitness and performance, and how that single number is not an accurate representation of a rider's fitness for racing. Many aspects go into a rider's fitness and even more aspects go into being successful as a racer, it is naive to think just a single number could accurately simplify all of these.

What is FTP?

FTP, short for Functional Threshold Power, is ubiquitous for estimating a rider's aerobic fitness. It has become so universal that almost all power-based training zones are determined as a percentage of FTP. Initially developed by Dr. Andrew Coggan, the co-author of Training and Racing with a Power Meter, it is meant to approximate the second lactate threshold, hence the name.

The second lactate threshold, often shortened to just the "lactate threshold", is the point where the rate of lactate production by muscle cells is roughly equal to or just higher than the rate at which your body can reutilize and clear that lactate. While FTP is generally regarded as the 'hardest steady state intensity you can ride for an hour', the physiology isn't binary like this. Some people may be able to only ride at their lactate threshold for 20 minutes while others may be able to do it for hours on end.

What is an FTP Test?

The 20-Minute Gold Standard

The most popular FTP test there is the 20-minute all-out time trial. If you have trained with a power meter for any period, you have probably done this test. You ride as hard as you can for 20 minutes, then you take 95% of whatever that power was and that is your FTP.

This is the first mistake riders make with this testing method, it isn't just riding as hard as you can for 20 minutes. The official protocol is to also do an all-out 5-minute effort just before the 20-minute test to try and minimize the contribution from your higher power-producing but less efficient muscle fibers. If you have ever done a 20-minute test without the prior 5-minute effort, you may have noticed that the first 5-ish minutes feel fairly comfortable despite your power being high, but soon your power starts to drop. This higher power at the start is you utilizing these muscle fibers and once they become fatigued they can not produce that power anymore so you are left with just your weaker, aerobic fibers to do the work.

Incorporating the max 5-minute effort at the start will minimize the contributions from these anaerobic muscle fibers and leave you with a more accurate representation of your aerobic fitness. That is the goal of FTP after all, to measure your maximum aerobic fitness without significant contributions from the anaerobic system.

The 2 x 8-Minute Test

A less common FTP testing method is the 2 x 8-minute test developed by Carmichael Training Systems. This test is simply an 8-minute all-out effort followed by a 10-minute recovery period and a second all-out 8-minute effort. This test calculates FTP as 90% of whatever the highest average 8-minute effort was.

This test is likely to overestimate your FTP and CTS clearly states this as a con of this testing method. Since this test takes the highest average power and not always the second effort's power, there is no way to minimize the contribution from the more anaerobic muscle fibers. You are more likely to produce higher power on the first effort since all of those muscle fibers are warmed up, not fatigued, and ready to go.

With this test being shorter in duration and not having a method to pre-exhaust the anaerobic muscle fibers, the results from this test will include a significant amount of anaerobic contribution. This contribution will likely lead to a higher calculated FTP than in reality, especially if you are a rider who has a higher proportion of those muscle fibers.

The Ramp Test

There are a handful of different ramp test protocols but I will be referring to the 20-watt/ minute one that can be found on Zwift. This test is arguably the most popular one out there because of Zwift, but many physiology labs use different power increases and step durations based on the specific goals of the test. This is a graded exercise test meaning you increase the intensity by a set amount (20 watts) after a certain amount of time (1 minute) until you can not go any longer. The FTP calculation for this test is a bit more complex:

MAP = (power of last completed stage) + ((% of uncompleted stage / 100) * (power of uncompleted stage))

FTP = MAP * 0.75

This test has the advantage that it requires no pacing strategy, so you can get an accurate result on your first test, and subsequent tests don't improve by simply getting better at the test. It also has the advantage that it is so short and only hard for such a small amount of time, you could realistically do this test before a set of intervals or regularly throughout a training block without risking overtraining, the same can not be said for the other two tests. However, because of how short this test is and the fact that it goes to absolute failure through the anaerobic intensity ranges, there will be significant anaerobic contributions which could lead to an overestimated FTP.

How is FTP Flawed?

While the concept of FTP itself isn't inherently flawed, the way most people use it can be. The process of using a series of repeatable tests to measure changes in fitness is perfect, without it, we would not be able to gauge how effective our training is. However, the issue comes when we try to extrapolate this value across a wide range of intensity levels and therefore physiological processes. Comparing your own result with a previous or future result is a great way to measure progress, but it is not necessarily specific and it can lead to inaccurate training data.

Overestimation of Zones

One of the cons for nearly every FTP testing format is that it is much more likely to overestimate your FTP than it is to underestimate it. While this isn't an issue for keeping track of progress if you are consistent with the testing format you use, it can lead to overestimating your training intensity zones too. If one of these tests overestimates your FTP by 10%, that means all of your zones are going to be too hard by 10%. This could mean your endurance rides are more like tempo and causing more fatigue than you are expecting, or it could mean you are unable to complete workouts entirely.

The Test Becomes the Goal

I know of several riders, myself included, who have had phenomenal 20-minute FTP test results, but get destroyed in real races that they are training for. For me, I was so focused on raising the amount of power I could do for 20 minutes that all of my higher-intensity performance suffered significantly. I told myself I was training for races but in reality, I was just training to get a higher number on my next FTP test.

A few years ago, I was able to do 340 watts for 20 minutes, but this year I am only able to do just a little over 300. However, back then, I neglected all of the aerobic capacity, anaerobic capacity, and sprint aspects of my training. Then I could only do about 380 watts for 5 minutes (119% of FTP) and barely ever broke 1000 watts in a sprint. Now with an FTP over 10% lower, I can still do 380 watts for 5 minutes (131% of FTP) and I have no problem breaking 1200 watts in a sprint whenever I want.

Remember that FTP is just a tool to be used in a larger toolbox of many other tools. Don't become consumed in testing better on FTP tests if it means you become even worse at racing.


My biggest issue with FTP tests is how unspecific they are. The gold standard 20-minute test doesn't accurately reflect the reality of racing at all. There are rarely any times in any format of racing where you will be riding at a constant, steady but maximal intensity for 20 minutes straight. You may think this happens in time trials or road breakaways, but it is always faster and more efficient to use more energy to push harder on the slower sections (climbs and turns) but recover a little on the faster sections (downhills and tailwinds). For this reason, I like to develop athlete and goal-specific tests that can more accurately predict the performance of their specific event. This is where having concurrent linear goals can be beneficial. If you'd like to read more about how to do that, check out our blog on setting and achieving goals.

What to do instead?

My best piece of advice when it comes to tracking changes in fitness is to develop your own specific test that is similar to your goal. If you are training for your first century, a ramp test will probably not accurately tell you how prepared you are to ride 100 miles. It can certainly show you progress over time, but that data doesn't directly translate to your goal. Maybe an athlete training for their first century would do a different test to determine how strong their endurance level fitness is. A great test for this is the talk test.

Talk test

This is also a graded exercise test just like the ramp test, but this is sub-maximal. This simply means you never push yourself deep, in fact, the hardest you will go on this test is probably easier than the warmup for your ramp test. It is so easy and relatively quick that you could realistically do this test at the beginning of each of your endurance rides as a warmup and collect data.

The talk test is best done on the trainer with ERG mode on so it can be precisely controlled, but it can be done outside too. Perform this test by staying seated the entire time and staying as still as possible to keep your heart rate stable. Starting with a 10-minute block at a power roughly 50% of your FTP (or the kind of power you would do on a super easy recovery ride), perform 5-minute block progressions increasing the power by 10 watts each time. in the final minute of each block, count out loud quickly and take note of how far you make it. Once you can't make it as far, you are done with the test.

Take note of what power you stopped at, your heart rate at that point, the RPE at that point, and how many fewer numbers you counted. These are all useful metrics to compare to the next test. If you are making aerobic fitness progress you will notice that the power you can hold at the end of the test will go up. This means for the same RPE, roughly the same heart rate, and the same breathing rate, you will be doing more power. This means you are more efficient aerobically!

Terrain-Specific Efforts

You can do specific tests to measure increases in your fitness and speed. This can be through testing your performance on a climb, section of trail, or other feature that you can easily track changes in. If performing shorter climbs at max effort is important for achieving your goal, you could find a 2-minute climb that you could do a max effort on at the end of each training block to track your progress. This is simple and specific to your goal, by getting better at short climbs you will test better on this short climb.

Another example could be if your goal is to decrease your total time to complete a 10-mile long section of trail, your test could be to regularly do a 2-mile section of it and see how long it takes you each time you perform it. This example goal is relatively short so you could even try for a full effort on the trail at the end of each macrocycle to be as specific as possible.

Need more help defining your goals and figuring out what the best way to do progress testing for yourself is? Check out our Beginner's Handbook to learn all of this and more!


Ride further, faster, and longer!

Join our weekly newsletter and elevate your training and racing with exclusive tips

Join now and get three FREE training plans

Get three TrainingPeaks plans for mountain bike, road, and Zwift! These plans include workouts and content you won't find on other coach's paid plans, for totally FREE!

  • Facebook
  • Twitter
  • Instagram
  • LinkedIn
bottom of page