HP vs Cadence curve for bikers

So there is the never ending debate about what cadence people should be biking at. But it should be fairly easy to determine this scientifically.

In performance automotive (and really all of automotive), we dyno cars and measure the torque and horsepower with respect to engine speed. The two are related of course, but I never see human power related to the cadence.

It would make sense to see where your torque output (force) on the pedals is most efficient based off a graph on torque vs cadence. I doubt it’s a flat curve… and likely different for everyone.

The big question is how do you test for power/torque at different cadence? Do the ramp test locked into a specific foot speed and repeat a dozen times? Or do a sweep from day 60rpm to 130rpm? You would obviously need a load bearing trainer to do such a test.

Or has this data already been determined and I just don’t know about it?

I think most people will find their own natural cadence that suits them best. You tend to see the majority spinning at around 90rpm. Spinning at different cadences are effective in training different areas, but I think that sweetspot of around 90 for most is he most effective/efficient.

+1 @AndyGajda…there is no one cadence. It’s a moving target and different for each rider based on fitness, fit, biomechanics, terrain…My opinion is it’s almost pointless to worry about it. Want more power? Push harder. Everything else works itself out.

1 Like

That’s not very logical. Why do we do ftp testing when it changes by how you feel in the morning and if you had spaghetti or a steak the night before…

It sounds like no one has given any (scientific) thought to it other than “it’s how it was, so that’s the way it is”.

What if testing for optimal cadence finds your 5-8 rpm off where you should be, and you can maintain power output 15% better?

Jan Ulrick was 60rpm and lance was 100 going up hills. That’s a huge variance, and if the common logic is for everyone to target 90, I bet a huge number of people are not where they should be.

It’s the same thing with force vs crank angle graphs. All these bikes with power meters and I haven’t seen a single one showing the distribution of where the power can be applied and how different pedaling exercises can improve it.

1 Like

What cadence you can sustain is down to technique. Higher cadence means lower torque at the same power, by definition. The higher torque you are putting out the quicker you are going to crack because you’re using anaerobic resources, which are largely finite, over aerobic resources, which are close to inexhaustible.

There isn’t a ‘magic point on the torque curve’ that you can look at. You just spin faster then monitor your heart rate and if you’re bouncing. If you’re bouncing or your heart rate starts climbing you’re going too fast. If not you can do that or higher.

Humans aren’t cars.

1 Like

Not saying there isn’t a cadence that is better or worse than another for any of the various conditions we find ourselves in. Just saying for any individual the best cadence changes during the ride, terrain due to many factors. The ideal cadence varies from rider to rider for many more reasons.

  • ^ This^

I get what the OP is asking. But we are not able to do the same performance in such a wide range as the modern combustion engine in a race car.

We are not trying to find the “power band” and maintain HP or Torque in the way they do with race cars. There it is a goal to shift and accelerate to keep the engine at optimum performance output, peak power and torque to travel as fast as possible on the track.

We kind of have a parallel with BMX and Velodrome racing. Those are massive power output at shorter duration with emphasis on pure speed for the most part. But when you look at the broader demand and use in real Endurance cycling, the parallel to mechanical engines breaks down.

We are more looking for a blend of power AND duration. Unlike an engine that is essentially unlimited on range (besides fuel), our human bodies are largely impacted by power over time. Add to that fueling as a complexity in reality, but even if you ignore that, you still have no real comparison.

Cadence can and does matter for power, as well as duration. Studies have been performed and there are good rules of thumb about what works for particular power outputs, and then add consideration for effort duration.

We may not have pure dyno-like numbers for each rider, but the opportunity starts with these decent recommendations. Then, each rider should be experimenting with their particular body, power, and event demands to find what range of cadence works for them.

Add to this that any particular event is likely to need a range of cadences for successful completion. You may be able to perform in a road race at mid level cadence, but if you lack high cadence for a sprint, you could be left hanging at the back. There is no single right answer for any rider or event.

The moral in my eyes is learn the basic recommendations and do some serious experimentation through your training and racing to find what works for YOU.

The main parameter missing in the equation when human beings (or other living things) get thrown into the power vs rpm curve comparison with engines is time. An engine can deliver the same power as long as you provide it with whatever powers it - fuel, steam, electrical power, uranium. So “how long you can sustain it” is part of the deal. You would likely find very different curves for 30-second sprints vs 1 hour efforts.

David Gordon Wilson"s “Bicycling Science” has such a curve - max power vs rpm - amusingly, it comes from a study comparing power achievable in forward vs backward pedaling motions, which study found that backward pedaling allows more power generation, but under conditions that remain to be reviewed. Wilson’s comments imply that the study was done with a single person in a single configuration.

Wilson is a must-read for cycling science geeks, by the way. It’s dated, but there’s not much out there that brings together all the science related to human-powered machines.

1 Like


Found the source:

Ramondo Spinnetti. Known for his weird backward-pedaled human powered vehicules (calling them “bikes” would be quite a stretch).


this graph is EXACTLY what im looking for. it shows the varience in power output with respect to cadence. and i totally get that every individual is different… which is part of the line of questioning i have is, why isnt there a test for this.

but using the graph data, it shows a 30 watt fall off with ±15rpm. to me that shows that it is basically critical to keep your leg speed where your peak efficiency is (ironic, that its exactly like cars). This is where bike gear selection can be important. if you know where your peak is, and you know the relative range where you are still ±5-10% of that peak efficiency, you can then know with a great deal of accuracy the gear ratios you should be using.

if we assume for a second that a persons peak is at 90rpm, using that graph, and give them a ±10 rpm range to operate at, we can see at 100rpm, they are around 155watt, and at 80rpm they are 175watt. the area under that curve is less then using 85watts as the center point. we are talking significant differences in efficiency/power output. 12% is far more of a difference then worrying about what position your hands are in for a TT… or what shoe cover has the least amount of drag…

and thanks for the book recommendation. im going to give it a read!

Gain of salt warning here (and why I posted a link to Spinnetti’s original article): it’s a n=1 sample, and note that the max power he outputs is below 200W, on a max effort. So he may be anything but representative of a well-trained cyclist. His point was to demonstrate that backwards pedaling is more efficient. Yes, I know.

And again, read the other posters: there is a huge difference between what can be efficient for 30 seconds vs 30 minutes; training can change you peak cadence; and the peak certainly varies widely between individuals.

1 Like

yeah i get that his test is to be taken with a grain of salt. it also doesnt really have a lot of data points, but it is something, rather then assumption.

So back to the original point, as you just mentioned again, the peak efficiency can vary widely between individuals, why do we recommend 90rpm to everyone? and if that peak changes vs the load, why isnt there a chart/test to determine this?

@mcneese.chad, you mentioned studies have been performed for cadence as it relates to power output. do you have links or articles i can find?

and somewhat related, do you know the update frequency of some of the powermeters on bikes over bluetooth or ant+? are we talking 10hz, 100hz, 1hz? if its relatively fast, say 50+hz, we can get a graph of pedal force vs cadance every 10 crank degrees. i would assume the higher the RPM of the legs, the more the applied force on the pedals shrinks in duration (even normalized over the period), which would correlate to the drop in efficiency at higher RPM.

1 Like

I have seen and skimmed various articles over the years. All I can offer right now are some quick Google search results that may or may not be the exact ones I have seen in the past.


The background of higher cadence= better is that you shift from muscular endurance to cadiovascular endurance - so again it’s a “how long can you do this” type of thing.

There is data in Wilson on force amplitude and direction vs angular position, and a lot of powermeters measure the power vs pedal position as well (the latter is different from the former, as it does not concern itself with unproductive forces).

1 Like

That first study and its conclusion that older cyclists have an efficiency peak at 40rpm = yeah sure, I can just see myself doing Kaweah at 40 rpm… as long as there is a replacement set of knees waiting for me after, no problem…


LOL, as mentioned, I didn’t read any of these right now. Just quick search results.

I would have to review the specifics of that study to see how and why they tested. It’s entirely possible it is founded on a silly thing like pedaling backwards as in the other study mentioned above.

1 Like

I think it’s generally accepted that a self-selected cadence tends to be the most efficient for THAT ATHLETE.

But you can re-train yourself over time so that a different cadence becomes your “natural” cadence.

FWIW, here’s my N=1 experiment.

I wanted to see how many watts I could produce in a max sprint one day (disclaimer: I don’t even have a sprint, and for the most part, I don’t care to), so at several convenient uphill locations on my commute home, I gave ‘er hell (slight heck, actually). I got home and looked at the numbers, and the next day, I tried again. Thinking I had been in too tall a gear before, I tried a lower gear this time and got a roughly similar result. Curious, for the next few weeks, I kept trying (when I felt fresh enough), futzing with different gears and cadences each time. The result: between 70 and 110 (ish), my max power was broadly similar (I wasn’t concerned with how long I held the power, just the biggest number that appeared on my Wahoo).

I had been hoping that if only I found the perfect gear, I’d find 200 extra watts hiding in my legs. Alas, those watts are still hiding.

On the complete other end of the spectrum, shortly after taking up cycling, I was on a climb, and a steep pitch sucked my wind, so as the gradient eased, I dumped it into the 32 and tried to recover. But I didn’t. I was in my lowest gear, for gosh’s sake, and it wasn’t getting easier! My legs kept hurting, and my hr and breathing weren’t slowing! Spin it to win it, right?!?!

It occurred to me that maybe an upshift would help? I clicked up one, then two, and holy cow, it felt better, and my breath was coming back (speed constant, cadence decreased). So instead of my little legs flopping around at 100, they were spinning nicely around 80, and so much for spinning it all the time, regardless.

As much as I’d like the idea of one Cinderella-slipper cadence, it seems that we meatbags have different muscle types and metabolic pathways that mix and blend and do different things at different times, and depending on alotta things, the “right” cadence can be alotta different numbers.

Like Mike Tyson said “everyone has a plan, until they get hit”

I often think okay I am going to stay at 93-95 rpm, then the ride gets tough and I just stay with whatever cadence is working best for me!

In training I stay pretty steady tho