How important is HR vs power for aerobic base training?

I’ve decided to do a full base/build/specialty approach in training for a goal race in March. I’m doing a hybrid of SSBHV and TBHV. I’m picking sessions from TBHV on Fri/Sat/Sun that have the same TSS as the scheduled SSBHV rides to achieve this. I’m finding that my HR is often in the middle of zone 3 at the top of power zone 2.

Is power the key determinant of aerobic adaptation on long, steady rides? Or should I dial it back a bit to keep HR lower in order to achieve the aerobic adaptations intended from these rides? I’m much less concerned with how my HR behaves at higher power output because I have always revved high and it is what it is, but I can’t help but wonder am I ‘aerobic enough’ for the purpose at the lower end?

The TR zones are all about power, and disregard HR, so if you want to follow the TR plans, ignore HR.

In more general notes, I don’t think you can determine how ‘aerobic’ you are from just your HR (there are long in-depth discussions on here about this whole topic). If you want to track HR for the long rides, what you’re looking for is that HR for a given power goes down with improved fitness. You can also track ‘HR decoupling’ for long rides. is great for that!


I’m gonna give you some bro science right now. So take it with a grain of salt.

I am in the same boat in terms of where I sit in my periodization: aerobic base building. My main goal for September and October is to accumulate as much power z2 as I can.

However… I often blend in some Power z3 tempo into the mix and thus my HR ticks up a bit more into HR z3. I’m comfortable doing a lot of that type of training (high z2 / low z3) and find that I derive a lot of training adaptations from consistently hitting they zone. This isn’t just my opinion either. I’ve heard a lot of other endurance coaches talking about prescribing a lot of work in that zone because it is almost purely aerobic.

One thing I’ve done this week to change it up was to go out and maintain a low cadence (65rpm or so) to tax more of muscular endurance. This keeps the HR low and the torque high (which benefits strength, biomechanics, pedaling efficiency, etc.)

Lastly (and somewhat unrelated, but I think it’s important), as scientific as training is, there is no body of research that will tell you how you’re feeling. I’m not a proponent of Whoop and other feedback devices. We ought to be able to know our OWN bodies. This season, I’ve learned to listen to my body. To the point where I’ve literally gotten up early to ride, felt like total crap, and decided to go back to bed (as extra rest was the better use of my time). If you’re feeling good, and your HR is just a little higher than you’d expect, don’t sweat it. But if there are other factors (HR is way high OR way low, you feel like crap or just tired), dial it back.


I’m going to assume first that you have an accurate idea of what YOUR maximum heart rate is, and are not just using the age-based default.

Probably the more relevant measure of aerobic adaptation is Aerobic Decoupling - or basically the extent to which HR rises over course of a long ride while holding power steady.

This is also something I prescribe heavily to. Watch the trend in your HR with the power your putting out over time during intervals. A high sweet spot and threshold effort interval that is 5, 10, 15min long will definitely find a leveling point. It’s an aerobic effort, so your aerobic system should get into a fairly decent steady state.

If your HR is constantly creeping up during steady state efforts, then it’s usually a good indication that your FTP is set too high and you should dial back the power to find the steady state needed to develop that physiological system.

Of course, there will be decoupling, we all fatigue. But watching this trend during workouts with medium length interval work is a good way of knowing you’ve got power, HR, and RPE aligned.

Edit: just as an example. This is my last interval from this morning’s workout Abbey. You can see that even at the end of the workout when I’m most fatigued, my HR achieved a steady state (I added a straight dotted line to help illustrate) and remained level. This is a good indication that my FTP and the workload prescribed in the workout are aligned with objective of improving my aerobic capability.

1 Like

Just want to ditto the decoupling comments, as an example of a really long ride. I rode 100 miles on Saturday at an endurance pace and experienced only 4.1% decoupling (under 5% on steady intervals is considered good).

I don’t pay too much attention to HR being in z3 during a z2 ride or anything, because my basement’s temp/humidity can be a factor in determining where it is on a given day. I wouldn’t stress trying to adapt to a HR vs staying at the recommended power for a workout.


Thanks for the feedback guys. I’d read other threads but found one very much in favour of HR, one very much in favour of power, and several that had no consensus, so I wanted to see the latest thoughts. And I’m glad I did because it has really directed me towards focusing on decoupling for these rides, which I’m going to do now (including supporting requests for it to be added as a chart in training peaks so I don’t have to track it manually).

I’m also looking more carefully at EF at the same time. These are metrics I haven’t paid much attention to before. I’ve struggled a bit since my last FTP bump right before starting my base training with TR and have probably been a bit too focused on wanting to get my FTP up even though I know it’s only one data point. I still want to get it up, but have a better appreciation now for how much improvement I can be making without FTP increases.

Yesterday I did Martin Hill and stayed under 5% until about 2:15 in. In my head I had it that it was a 2:30 session and the wheels came off when I realised I still had 30 minutes with a higher block still to come. Last week I was good for about 2 hours of Hawk Mountain and just over 5% for all of Gibbs. Doing Koip today so hoping for improvement there.

And with that, thanks again and I’d better get myself on the bike!

1 Like

Pretty good bro science :fist:t2:

Endurance on the flats, tempo on the hills. Nothing wrong with that. just got a new page for those interested in power vs heart rate. The power vs HR chart shows changes in fitness very clearly. As you get fitter your curve will move to the right (more power for the same HR). Very clear fitness indicator with no testing at all!


Great work, David! Looks like you’ve had a nice little bump to your fitness. Very clear indicator indeed.

My chart has a bit more noise:

You too! Unfortunately I have slacked off a bit since then and am not that fit currently.

The graph gets very noisy once you reach threshold as the linear power/HR relationship breaks down + there is less data. I left that in because it shows very clearly where your threshold is for power and HR.

Stick to the power. HR can be affected by many factors so I wouldn’t be worried about it much.

This is nice! (Also that I’ve finally found a graph that says I’m getting fitter, haha). Where do the data points come from, the PD curve? And are they average over the timefrme specified?

what’s the methodology for putting together the chart? Like, it doesn’t just give you the highest power for a given HR, right? If so then going and doing something like sprints seems like could skew it as you might be at like 1300 watts while still at 130 bpm as your hr catches up, then hr peaks after the interval.

Tx. The lag adjusted power vs HR data for each moving minute of each ride (as shown on the ride decoupling charts) is bucketed into 5w buckets (e.g. 220w, 142bpm, 20 minutes). The curve for a time period is the average of these buckets for all rides weighted by the number of minutes in the bucket. So if one ride has 220-224w 140bpm for 10 mins and another 220-224w 150bpm for 5 mins the then bpm for that bucket = (140 * 10 + 150 * 5) / 15 = 143bpm.

@devolikewhoa83 breaks your ride down into 1 minute segments, computes the HR lag then calculates the avg power and lag adjusted HR for each. Still you do get anomalies on the high power side e.g. if you do a 60s flat out effort. The chart is most interesting from a bit above threshold and lower.

It’s really interesting how nicely the chart shows threshold power.


Thanks, thats a lot of computation!

Yeah I’m also intrigued by the ‘threshold kink’. It aligns so well with what I think my threshold is/was that I’m wondering if the ftp setting is somehow influencing the data? Weirdly for a year where I was less fit, the kink is not only earlier, but also at higher HR. I wouldn’t have thought threshold HR would come down?

There is something a little odd with the comparison marker when you go above threshold, but I don’t think you can do anything about it. Because the line goes approx horizontal, the marker sometimes picks power values quite far away from the one you’re looking at for comparison. Maybe it could also show the difference in HR for a given power? (Or an option to switch).

Yes the marker does do strange things. It find the curve with the lowest HR for the given power (interpolating between points) and then uses that HR to find the corresponding power on the other curves. Supposedly.

An option to show HR difference instead of power is a good idea.

Re threshold HR coming down I agree I would also expect it to go up a bit. Maybe someone can explain that?

But that makes me think though, if you’re having a crappy HR day (say your RHR is 10bpm over) and still want to get an easy ride in… if Pettit gets your heart going too hard, as an example, you might want to reduce the intensity to keep the HR in check no?

Otherwise you might be going past the zone that keeps you in endurance mode and strain your body too much.

But that’s just a thought about how i’d approach it, aka my $0.02

1 Like

I don’t know, I always stick to the power regardless if my hr is elevated or not.

I think that you need to look at both of them in perspective. Go by the power numbers and use the HR to see the improvements. If you ride at 200w for 2 hours at 140bpm and after base training you can ride at 200w for 2 hours at 135bpm than you know it’s working.

Also look at the aerobic decoupling aspect too.