Do You Know If You Are Training and Riding To An Accurate FTP?

Many masters cyclists treat their FTP number like it's set in stone.

They do the test, they get the number and then that number runs their training zones for the next three or four months without any further thought about it. Every session, every interval, every easy ride… all calibrated off that one result from that one day.

But… is that single test and result from one point in time truly accurate for you?

What FTP Actually Is (And What It Isn't)

FTP was originally defined by Hunter Allen and Andrew Coggan as the highest average power you can sustain for approximately 60 minutes. In practice, most cyclists estimate it via a 20-minute test and take 95% of the average power. It's become the go-to metric in cycling from which entire training plans and workouts are then built from.

But here's the thing about FTP... it is not the same as your true physiological threshold. Research has consistently shown that while FTP correlates with lactate landmarks, it is not equivalent to them, with variations as wide as 45-50 watts above or below depending on the individual and the lactate method used [1, 2]. Read that again... 45-50W either side...!

A better true physiological marker is CP (W') which, compared to FTP, is usually higher in trained cyclists.[3] But, CP is more physically taxing and mentally gruelling to test. So, most people default to the standard FTP test. And as such, if you're basing your entire training plan on one FTP test, you are opening yourself up for variable training outcomes.

This doesn't mean FTP test results are useless and it doesn't mean you can't use it. The key is to understand that it is a practical proxy, not a physiological absolute and you need to account for the practical limitations.

A Single Test Day Is Fraught With Error

Taking the result from one single FTP test from one single day and then dictating 3 months of training around that can result is hugely ineffective training and outcomes. But, where FTP testing can shine is in repeatability...

Research on FTP test reproducibility in trained cyclists shows excellent test-retest reliability, when conditions are well controlled [4]. But if those conditions are not controlled, individual test results still vary from day to day based on what conditions surround you on test day....

Sleep restriction alone has been shown to reduce cycling performance and push perceived exertion up for the same power output [5]. Time of day matters… evening cycling performance consistently outperforms morning performance in time trial studies [6]... the research says "consistently", but I find some people are better morning people. I am!

Hydration, residual fatigue from the previous few days, life stress and even what you ate the day before all impact FTP testing. As such, your FTP test isn't just measuring your fitness in isolation. It's measuring your fitness on that particular day, compounded by everything else that is going on for you at that time.

So, again, the key is to understand the practical limitations and that the more FTP testing and variable controlling you do, the more accurate your metric gets and the more accurately you can refine your training zones.

I certainly don't mean testing every week... but ensuring that you test regularly enough and under controlled circumstance in order to accurately refine your metric.

Why This Matters More After 40

Recovery status has a bigger say in test-day performance for masters athletes than it does for younger riders. If you test on a day where you have residual fatigue, or you slept badly, or work has been relentless… your result can come be well below your genuine capability.

And if you then base your training off that lower number for the next few months, your zones end up too easy. Your "threshold" work isn't actually at threshold. Your VO2 efforts don't create enough stimulus. And you may start to wonder why you're not progressing when the answer isn't your training… it's the test you did and the number you're basing it all on.

It works the other way too. If you test on a day you're super fresh, or riding on adrenaline in a group test environment or just have a day "out of the bag", you'll get an inflated result. Take that number and use if for training make you feel like you're going to get a great result after 6, 8, 12 weeks...

But, in reality, your zones are too hard. Every "easy" ride becomes a grey zone for your personal metrics. Every interval makes you more fatigued than it should. And, after 40, when the buffer for training mistakes gets smaller and recovery takes longer, both of these errors compound very quickly, which will only service to undermine weeks and months of effort.

A Smarter Approach To Testing

So, how do you make it smarter?

What matters is tracking trends over time, not relying on one single test and one single number. This means that your numbers and zones are roughly but consistently right rather than tied to one single exact number from one single test.

Here are the things to keep in mind:

Standardising your test conditions. Do them at the same time of day. Same warm-up. Same fresh state. Same protocol. Every time. This way you're comparing like with like.... Apples to apples.

Tracking trends across multiple tests. One test tells you what happened that day. Four tests over a 4-6 weeks tells you what's actually happening with your fitness. Doing this also acts a proxy for interval work!

Factoring in context. A big PB on a day you felt terrible compared to a lower "modest" number on a day you felt fresh and rested is important information!

Using multiple data points. Ride data from your training and not just from formal testing is hugely valuable input. Modern platforms can estimate your effective threshold from weeks of accumulated ride data, so use this to inform your actual reality rather than relying on one single all-out test.

How often to retest, how much weight to give recent results versus older ones and how to build zones that actually reflect your true capability is where it gets more individual. But it's worth getting it right!

If you're starting out, this means doing a number of FTP tests in succession over 2 - 3 weeks. It will be exhausting if you have not fitness base, so factor that in. Perhaps you need to skip FTP and work on your foundations first! But if you have a solid foundation, then a few tests will give you trends from which you can zero in on your range.

Conclusion

FTP testing is a genuinely useful tool, but treating a single test result as your "unquestionable truth" opens you up to problems. Take a more critical look at your own testing history… were you fresh? Tired? Stressed? Well slept?

Over time, building a smarter relationship with your numbers that is based on trends and context rather than one single day is what lets your training zones actually reflect where you're at. And when your zones are right, your training starts working the way it should.

Frequently Asked Questions

How often should I retest my FTP?

Most masters cyclists benefit from retesting every six to eight weeks during a structured training block... provided you start from a solid foundation. What matters more than how often you test is that you test under similar and favourable conditions each time, so your results are genuinely comparable.

Does time of day affect my FTP test result?

Yes. Research consistently shows meaningful differences in cycling performance based on time of day, generally trending toward better output later in the day. If you're starting out, start with end of day but understand this is highly individual and you may need to refine. Once you discover what works for you, stick to that! If you test at different times each time you retest, your comparisons will not be accurate… so keep it consistent.

Should I trust my smart trainer's auto-detected FTP?

It doesn't replace a proper structured test. But it can be a useful data point. Most auto-FTP algorithms use short peak efforts or ride data without accounting for your recovery state and other individual factors on that particular day, so treat it as one input among several rather than your gold standard.

What about critical power (CP) or the ramp test as alternatives?

Both have their place. Critical power is arguably more physiologically grounded than FTP but requires multiple test efforts to calculate, is more taxing and mentally gruelling. The ramp test is quick and accessible but can skew results depending on rider type and it often overestimates FTP in riders with strong anaerobic capacity and underestimates it in diesel-type endurance riders. What matters most is understanding the limitations, picking a protocol and sticking with it consistently. Don't switch between methods… this will just muddy your trend data.

References

1. Jeffries, O., Simmons, R., Patterson, S. D., & Waldron, M. (2021). Functional threshold power is not equivalent to lactate parameters in trained cyclists. Journal of Strength and Conditioning Research, 35(10), 2790-2794. https://pubmed.ncbi.nlm.nih.gov/31269000/

2. Sitko, S., Cirer-Sastre, R., Corbi, F., & López-Laval, I. (2022). Functional threshold power as an alternative to lactate thresholds in road cycling. Journal of Strength and Conditioning Research, 36(11), 3179-3183. https://pubmed.ncbi.nlm.nih.gov/34127613/

3. Karsten, B., Petrigna, L., Klose, A., Bianco, A., Townsend, N., & Triska, C. (2021). Relationship between the critical power test and a 20-min functional threshold power test in cycling. Frontiers in Physiology, 11, 613151. https://doi.org/10.3389/fphys.2020.613151

4. Borszcz, F. K., Tramontin, A. F., Bossi, A. H., Carminatti, L. J., & Costa, V. P. (2025). The reliability and construct validity of the functional threshold power test in trained cyclists. Journal of Science and Cycling, 14(1), Article 6. https://doi.org/10.28985/1425.jsc.06

5. Dean, C., et al. (2023). Sleep restriction between consecutive days of exercise impairs sprint and endurance cycling performance. Journal of Sleep Research. https://onlinelibrary.wiley.com/doi/10.1111/jsr.13857

6. Souissi, N., et al. (2014). Effect of time of day on performance, hormonal and metabolic response during a 1000-m cycling time trial. PLOS ONE. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4188634/