Mountain biking TSS - some ponderings

Having compared my hrTSS and power-based TSS scores for indoor rides over a number of weeks, I had previously thought that hrTSS was a good enough estimation for use when power was not available (all outdoor rides for me). However, yesterday I did a 2 hr 40 min mountain bike ride and my calculated hrTSS was 233. This is clearly way too high and would have suggested that I was flat out for the entire ride (I wasn’t). Using the TR TSS estimator based on RPE gave a much lower figure of around 130, which I felt was a touch low but much more believable. Increasing the RPE by 1 gave me a value which I was happy with.

There are a number of reasons that I can think of that the hrTSS would be so high for mountain biking:

  1. Most of the time I was coasting was on fast downhills and heart rate would have stayed high due to the adrenaline.

  2. Periods of coasting were generally interspersed with short periods of sprinting, keeping heart rate elevated.

  3. Heart rate would have often been higher for a given power output compared to road riding due to the amount of upper body work required on technical trails.

This last point in particular got me questioning how valuable TSS really is when undertaking different types of riding or different sports. If I had a power meter on my mountain bike, wouldn’t the TSS score be vastly underestimated as it would only take into account the stress of pedalling and not the stress of the upper body requirements? But then, does that matter if you’re using TSS as a way of planning weekly training load? If your upper body is tired but your legs are fresh then you should still have a similar capacity for indoor training than if your upper body was fine, right? Or will any stress impact your capacity for training, irrespective of the part of the body that is stressed? If the latter, then maybe there should be different calculations for estimating off road riding TSS and road riding TSS.


hrTSS from TP?

It’s funny I had thought about posting up to say I find it reads a bit low for Mountain biking. Assuming the number it’s the value from TP - have you set your Threshold HR correctly?

1 Like

Yep, hrTSS from TP. Threshold heart rate and all zones set correctly using Joe Friel’s zones. On trainer rides there is very little difference between hrTSS and power TSS.

I literally just put a power meter on my MTB because I was tired of hrTSS lol.

I’m not sure what kind of downhills you have in your area, but in my area we have nothing that’s very sustained, so the upper body load is very minimal. After a really hard MTB workout, the fatigue from the upper body is gone within a day.

For me, I don’t care much about the upper body fatigue. I believe it does play a part in your overall fitness feel though. But typically, I don’t think I’ve ever done a hard MTB workout and woke up with fresh legs and fatigued upper body. Usually when my upper body is tired, so is my lower body (unless I crashed).

The trails in question are pretty technical (both downhill and some of the uphills). Towards the end of the ride I was still feeling like I had plenty left in my legs, but I was suffering from arm pump and my shoulders were tiring. This will largely be because I haven’t been mountain biking in a long time (in fact I hadn’t ridden outside at all since my second son was born in July), but it did make me think about how much this would add to TSS.

You might be right though. Perhaps the amount of additional stress is insignificant as a proportion of total stress and it just seems otherwise due to the fact I’m working different muscles from normal.

My experience says hrTSS needs to be interpreted thoughtfully to compare with powerTSS - for the same reasons the OP indicated.

Example: I did a 6 hr MTB race with a hrTSS of 603 - which would imply I was riding at threshold for 6 hours!

The RPE based approach is a good way to check hrTSS. If hrTSS looks too high, it probably is.

If your HR is high from downhills and your body is responding to the adrenaline by elevating the HR I would still say this counts as stress. TSS is a way to quantify your workout based on relative intensity and duration. It’s not a measurement of work like power. So even if you aren’t doing any “work” through the pedals while coasting, you are still stressing your body. Downhill MTB racers have to have a very high aerobic capacity for this reason. The original TSS concept was actually developed using HR-based training. It can of course be used with RPE, but to me this can be tricky. Especially because of the subjective nature of RPE. Like you pointed out with adrenaline being high, those types of things can warp our subjective perception of the actual stress and effort being imposed on our bodies. I can say from experience that I will finish a MTB ride with flowy downhills, feeling quite stoked and forget how hard I was working on the climb. I personally like to look at hrTSS and then compare that with my RPE score and more often than not, they are pretty similar anyways. If not, I will make a judgement call, rounding up rather than down.
It is important that you have your HR zones and HR-MAX correct, which you have said they are, but with such a high discrepancy, I would think maybe take a look at the zones once more.

It seems though, from what I’ve been hearing lately, the direction most top level athletes and coaches are going is actually RPE and maybe HRV as well, to keep track on overall training stress. This does require you to have some experience and be in tune with your body.


I’m pretty confident in my HR zones. They were determined during a fairly recent 20 minute FTP test, using Joe Friel’s formula to calculate FTHR and determine zones from that. Max HR I’m not sure of, but my understanding is that this isn’t required for hrTSS calculation.

The only way to get the RPE-based TR-calculated TSS to match the hrTSS is to set the RPE to 10! This was definitely not an all-out ride. My average HR for the ride was 155, which is bang in the middle of my Sub-Threshold zone. This fits with the RPE of 6 for the ride that I felt to be an appropriate choice using the TR RPE-based TSS estimation. I guess that the weighting in the hrTSS calculation really bumps up the score when there are peaks in HR above threshold, exacerbated by the fact that it will take a while for HR to drop even when intensity has decreased.

I wonder if selecting RPE in the TR tool based on average HR for the ride is a valid way of getting a decent estimation - I’ll pay attention to this in the future after MTB rides.

How long was the ride? I know duration is also a factor in using hrTSS and TSS. I would be interested to see how much time you spent in your zone 4 sub threshold compared to the other zones.
FTP = 100 TSS / hour. With hrTSS, 100 TSS would be 1 hour spent at LTHR.

But that is very odd that the RPE would only match hrTSS when set to 10. It would guess the issue is with TrainingPeaks in that case. In that link it does say hrTSS is problematic for intense, or hypervariable efforts due to the body’s limited response time with changes in intensity.

2.5 hr ride: 7.5 min Z1, 13 min Z2, 13 min Z3, 1 hr 2 min Z4, 38 min Z5a, 21 min Z5b, 6 min Z5c (all rounded to nearest minute)

hm so you spent over 2 hours in zone 4 and above. That type of HR effort would typically be associated with race pace. It makes sense to me that TP would give you a high hrTSS based on the HR zones you’ve entered.
Downhill MTB could be a unique sport in some ways and I’m not sure if they’ve looked into that specifically.
In any case that ride was stressful on your body and aerobic system to say the least. Not sure if the hrTSS can quantify it exactly. But at least you can manually override it with RPE.
Can you confirm that the highest 20 min average from that ride matches your LTHR?

1 Like

Touch higher: 166 vs 161

You might want to consider changing your LTHR to that especially if you see that often. It could mean that you tested lower than what your physiology is capable of. That might make hrTSS match a bit better. Maybe it’s omething to consider especially if you do a lot of those types of rides and see your HR up there for longer periods.

1 Like

I tried bumping up my LTHR to 166 and recalculated hrTSS for the ride. This brought it down to 203, so still much higher than the TR RPE-based estimate. Looking closer at my HR values from the ride, my peak 60 min HR was 161, which was what I had my LTHR set as despite this not being a threshold effort. However, my indoor rides don’t show large discrepancies between hrTSS and power TSS when hrTSS was calculated with a LTHR of 161, and hrTSS calculated for outdoor road rides seem sensible.

Given the above, I’ve come to the conclusion that my HR zones need to be set differently for road cycling and mountain biking. This makes sense - it’s well known that HR zones will be different for running and cycling due to the different muscle groups involved and the weight-bearing nature of running, and it follows that there will be a similar difference between mountain biking and road riding or indoor training.

Determining correct HR zones for mountain biking would be tricky due to the variable nature of the sport, so I think I’ll stick with hrTSS for outdoor road rides where I don’t have a power meter, and use the TR RPE-based approach for mountain biking. This should be a good enough estimate of TSS to evaluate whether any adjustments to my indoor training plan should be made.

Yeah I think RPE is the way to go then.

I found this study to be quite interesting

The conclusion was: “…mountain bike downhill races are conducted at high heart rates and levels of blood lactate as well as increased concentration of salivary cortisol as marker for psycho-physiological stress.”

So my only caution would be, not to underestimate the stress on your body during those rides and make sure you are recovering properly to avoid overload.


Found this thread and I’ve been pondering these same questions for a while now. My hunch, as well as talking with other competitive mountain bikers, is that TSS doesn’t really work for mountain biking in the sense that it does for road riding if you are using TR.

Anecdotally, a mountain bike ride with a stress score of 150 for a 2.0 hour ride is much easier, both in terms of body recovery and RPE (usually), than a 1.5 hour TR threshold type ride. I think it has to do with the constant stress TR has on your legs (and the build up of leg fatigue) vs. the many micro bursts you would have riding something like comfortably numb in Whistler. If you had a power meter on your mountain bike and did comfortably numb you might have somewhat similar times in the same zones because of the constant punchy rooty climbs that ride has (in terms of overall time spent in a high zones) but it’s micro bursts and not sustained effort like most TR workouts. The later being much harder on your legs in terms of stress,

I’ve seen better improvements in managing fatigue and recovery this year by decoupling TR ride TSS and outside mountain bike TSS. Intuitively I think they don’t relate well together to provide meaningful planning or as a training/recovery tool. I started looking at weekly goal stress/hours for each discipline and plan my week and month accordingly and I have been managing my recovery much better. YMMV.

1 Like

Anyone go back to hrTSS after using a power meter for MTB after finding that the power meter did not adequately measure TSS? About to pull a trigger on a PM for my XC bike but second guessing the need to buy one if all i am going to do is question whether the TSS score is accurate. Seems I may as well use hrTSs and RPE…

Who came up with hrTSS? Anybody know how it is calculated?

If your only concern is tracking TSS, I’m not sure that’s a strong enough reason to get a PM. I use one for pacing, post ride analysis, and performing outside workouts.

Plus it’s neat seeing the work you’re actually doing in real time time. Lots of micro burst 500+ W efforts on punchy trails. What’s more interesting is all the time spent coasting. TSS doesn’t account for the stress accumulated by tough descending.


On my road bike I have PM and measure TSS, but on my MTB I don’t have PM and I’m using hrTSS to get load from both trainings. I’m also getting hrTSS from rowing and running. I’m combining all results in both Elevate for Strava or TraingPeaks.