20min FTP Test Pacing Strategies

Original post in my Strava: https://www.strava.com/athletes/34093939/posts/12275269

So I wanted to experiment with different pacing strategies for a 20min FTP test and see if I can come up with some customized rule of thumb for myself.

We all have experienced at least one form of pacing strategy for a 20min FTP test. It is also well-known that many coaches, e.g., Chad Timmerman (TR) (https://www.trainerroad.com/blog/author/chad/), etc., prefer a Ramp test over a 20min FTP test for the very reason. Some, however, prefer a longer test to normalize the pacing mishaps as they believe a Ramp test, depending on the VO2Max performance of an athlete, overestimates the actual FTP (https://www.empiricalcycling.com/about.html).


  1. Experimenting with different pace strategies pertains to a 20min test
  2. Stress-test my body to these 5 tests in a row with no rest day in between. (almost accomplished this!)

Exhibition Loops, Toronto: 5x (https://www.strava.com/segments/26055573)

  • Each loop 2.4k
  • Total distance 12.2k
  • Estimated time per loop ~ 4.30 min
  • Estimated time per 5x loops ~ 21 min


  • Warmup: For all these tests, I warmed up for 15 min easy spinning and 2x1 min threshold followed by a single neutral loop around the route.
  • During the test: I tested out 5 different pacing strategies shown below. Time was not a limiting factor, but the distance was, and thus, depending on the average speed of the day, there are between 20:36 to 22:52. Factors impacted the average speed were: (i) avg power, (ii) wind and weather condition, (iii) riding position. I tried to be on the drop as much as I could and I think I almost hold it for 80% of the total time.
  • Cooldown: 20-30min of active recovery and easy spinning.

Experiment Limitations:

  1. Very very few data points
  2. Fatigue factor
  3. Rest and recovery including the nutrition consumed before and after the tests
  4. The effect of Weather condition on HR

Five data points are nothing but noise as oppose to something which can be inferred from, however, I feel my fourth pacing strategy (4th day) was more intuitive and natural for me to implement. I felt good and strong even after I finished it. Overall, on all 5 attempts, I did not reach my lactate threshold HR which is around 175-176 bpm and it shows that I have underperformed in all instances. That could be reasonable as I didn’t taper for any test, and these have been done back to back and accumulated fatigue could have played a role in the outcome.

Would be happy to hear your thoughts.

Very good post and interesting to see the results.
Obviously, the idea of a 20 minute test would be, to hold the same power for 20 minutes and be completely exhausted at the end. This is of course almost impossible to achieve.
My biggest limiting factor has always been “motivation”. I never have a lack of motivation to get off the couch and on the bike. But a 20 minute test is daunting and means 20 minutes of suffering.
If I don’t feel very good about and start out too eagerly, my results tend to be a lot lower than they can be. Every once in a while I have these “you don’t feel the chain”-days, and if that coincided with a 20 minute effort, I go all out, which usually results in a new FTP.
I believe, that a personal pacing strategy is favorable, so probably the best way to find that out is frequently putting yourself through similar efforts. As a timetrialist, I do a lot of work at threshold in the 15-25 minute area, so I have a very good starting point right there.

1 Like

For me I prefer a negative split. U se a spreadsheet with 4 columns each representing 5 minutes, with a sum average at the end. I put gradually increasing watts in the columns until the average comes out to what I feel I can/should achieve e.g. 305w, 310w, 315w, 320w = 312w. Of course on the last 3 mins onwards I start pushing hard, but overall and especially after 10 mins I’m judging my RPE being careful not to push too hard.


Interesting take. Thanks for sharing.

Right. Something similar to my 2nd and 3rd-day pace strategies :ok_hand:

1 Like

A workout I recently did for another group did a decent job estimating where to start a 20 minute test three days later.

  1. z1 warmup
  2. z2
  3. 10 min z4 (full gas)
  4. 5 min z1-z2
  5. 10 min z4 (full gas)

Take the average of steps 3-5 and start the first two minutes of the 20 minute test then adjust power as needed.

My average was 255 W for the test run steps 3-5 so I started there then went up to a target of 265 W the remaining 18 minutes. Ended up with average power of 265 and HR of 178 on the 20 min test, both were PR’s. Threw up a little after the 20 minute test and I really struggled at the end so I think I left it all out there. In the test run, I had traffic screw up the workout a little but I think it did a great job giving me a starting point to pace from. I was shooting for 275 in the 10 min sections, hit 271 and 273 for them which felt just a little too much for a 20 minute effort so for the 20 min test I thought I could possibly do 265.

1 Like

Interesting method @Trippy :ok_hand:

1 Like

I’ve done 5-6 20 min FTP tests in the past 2 years since I first started structured training. I’ve honed in on and used this strategy for the last few which has gone pretty well.

  • Have a realistic guess for what I think my FTP will actually be. Should be able to come up with an educated estimate based on current FTP and my feeling and performance of recent workouts.
  • Break the 20 minutes into four 5-minute buckets.
  • Aim for “target FTP” (guessed at above) minus about 8-10 W for the first 5 min, and increase the target power by 5 W every 5 minute interval. I sometimes adjust the bump up in power (more or less than 5 W), depending on how I think it’s progressing.
  • The last 5 minutes, I should know if I’m barely hanging on for dear life, if I think I could hold, or if I think I can push a little. Obviously I try to push a little if I think I can finish it out.
  • Through the entire test, every 1 minute, I switch between the higher and lower end of my natural cadence range. This usually means flipping back and forth by 1 gear. This is typically something like 1 min @ 93 rpm, 1 min @ 85 rpm. I don’t know how this impacts the relevance of the test or my physiology during the test, but it helps me mentally, even if it is only something to focus on for “1 minute at a time”.

No idea if that is a good or faulty strategy, but it has worked pretty well and at the end of the day, I try not to overthink it. I figure that if I’m relatively consistent in my approach, that even if my FTP result is off from my “real” (actual physiological, but not possible to measure) FTP by even 10-15 W, it should be productive and result in an FTP that will produce good training adaptations.


Thanks @tanzbodeli for the extensive description :ok_hand: Yes. I do tend to switch gears to alternate cadence, especially if I am getting tired and the more tired I get, this becomes more aggressive and distinct between the two. Say, high cadence at around @ 95-97 rpm and low at around @ 81-83.

FWIW in March I did a 20-min all-out effort with an attempt to over-pace at the beginning:

which gave 243W estimate. Then 9 days later verified with a 50 minute all-out effort:

Then 3 months later did a more evenly paced 20-min all-out effort:

and bumped up to 260W estimate.

Looking at 4 years of power data in WKO4/WKO5, my 20-min best is always around 95% of 50-70-min tests, so that multiplier works well for me.


Interesting thread @AmirJamez! I preferably do the 20 minute test of the ramp test, as I believe it gives me a more accurate ballpark of my FTP.

Basically I look at my power duration curve and see what I achieved last time I did the 20 minute test (as a reference point), and also the point where in that 15-20 minute (sometimes it is a little before that) range where my power starts to decline. Assuming I was feeling strong in my workouts leading up to the rest week, I then start my first 5 minute block at that wattage. If I am feeling good, I then try to up the wattage every 5 minutes. This seems to work best for me, and typically results in a pretty accurate assessment of my FTP.

1 Like

@bbarrera Good observation of your improvement!

@cnigro I can’t agree more with your pacing strategy. I guess a similar method gave me the best result out of those 5 attempts.

oof… tough week! I think your best power in test 4 most closely aligns to the negative split approach that I try to do below, which is even more telling considering it was towards the end of the week.

This is what I try to do for a 20min test or a shorter 8-10mile flat TT.
~5 mins at FTP
~10 mins at target wattage for test/TT (gradual ramp up to target; adjust up/down half way based on RPE)
~5 mins all out (ideally after middle 10min you’ve ramped up to max sustainable, so this is ideally just holding on; if you have enough to make a big jump in power you’ve held back too much)

I’ve made the mistake of starting out too hard too many times, so now I make sure to force myself to be conservative for first 25%. I may start to ramp up after 2-3mins if FTP is feeling easy at that point. As I’m sure you know based on your experiments, the worst feeling is getting 8-12mins into the test and realizing you’ve overshot and feel like you’re going to die.

1 Like

Cheers @harrington! Yes. Your approach is pretty close to what I tested in the 4th attempt and yes, I feel like I had more to give out since I was feeling pretty good after the completion of the test and I saw a 5-7w jump in avg power during the final lap (5/5). Perhaps, I should try to ramp up more during 25% - 75% of the test and see if it helps, this time for sure with sufficient tapering and rest lol

I’ve always tried a steady pace for the first 10, like a number that would put me right at, or a hair above my known/predicted FTP. Then try to negative split by just a hair for the next 4:30 to 5:30, depending on how I’m feeling. Then, presuming I’m saying to myself, and believing myself, “I got this” the last bit I try to overshoot appreciably, and this is usually the first time I’m being forced to change gears/cadences dramatically.

1 Like

Here’s a screenshot of the pacing spreadsheet I use for my 20 minute tests:

Each column is a different configuration of the test in five minute intervals, with row 5 being the average power of the whole 20 minutes. The final row is the FTP result.

It’s worked quite well for me and I now prefer the 20 minute test to the 8 minute or ramp test. This is all inside, but you could easily apply this logic to an outdoor test.

I also have found that relaxing my muscles during the test really helps me deal with the build up of discomfort and heaviness.

1 Like

Interesting @daturaman. Seems like all 6 of them are negative splits with different increasing rate. Pretty close to 300w :slight_smile:

Yes, I’ve managed to get to 303w (indoors) so far with this pacing.

1 Like