I read somewhere that steady cadence is a good thing but why is that? It is analogous to running long distance where it is a good idea to set a reasonable pace and stick to it? I would think that varying your cadence would give your muscles a chance to rest some. I question whether the original claim is actually true. I would think it would be more efficient to accelerate briskly then coast down, then repeat.

This "sawtooth" cadence question is interesting because with a car or motorcycle, you actually get better gas mileage if you accelerate briskly and then just coast down. For example, you are at a red light in pole position (first). The light then turns green. You then accelerate quickly up to the speed limit (say 50 MPH) and then let the car/motorcycle coast down to maybe 20 MPH and then repeat. This is actually more efficient (you will use less gasoline) than if you accelerated slowly and maintained the 50 mph. This has already been confirmed by many people that have tested it. It is also easy to confirm yourself, especially if you have a real time mileage display. It will be at maybe 10 MPG for a brief time while you accelerate then vary between 99 and and some lower number as you coast down.

Ok now to bikes. It seems to make sense that for a bike a similar thing should happen. Imagine a person pedaling an average bicycle 10 MPH on level ground with no wind. The amount of energy expended by the rider depends on several factors including what gear they are in. Imagine they are in the proper gear vs. much too tall (high) a gear vs. much too short (low) a gear. Problem is in the low gear energy is being wasting pedaling faster than is needed (for example 150 cadence). This is a "subpoint" I am trying to make related to the main point of "sawtooth" cadence I am working up to.

If I pedal 10 miles at 10 MPH at near constant cadence, that will be 1 hour straight on pedaling. How can my muscles ever get a rest? Compare that to pedaling up to 20 MPH quickly (briskly) and then coasting down to 3 MPH and then repeating. Here I use my muscles for a quick burst then spend most of my time coasting. Here my muscles get a good rest and the number of actual crank rotations is far less. 10 miles on my bike in my favorite middle gear (34/20=1.7) with my 26" diameter tires would require about 4500 crank turns. I suspect using the sawtooth cadence method would require far less, possible only 1/3rd or about 1500. It would be an interesting test to have someone make the trip using both techniques and report their level on fatigue afterwards.

Someone mentioned about pedaling then coasting a lot will make me look like a kid on a BMX bike. Well kids are generally lazy. Maybe they figured out it is easier to pedal and coast so that is why they do it.

Another analogy is at my workplace. Can I get more work done if I work hard but take several breaks or if I just work straight thru without breaks but I get more tired that way? For me, I need the breaks to stay productive.

Maybe for younger riders a steady cadence is better but I am wondering for the 50+ crowd if sawtooth is actually better. I suspect it is.

  • 7
    Have you read the answers to this question? bicycles.stackexchange.com/questions/12518/… In any event, if you examine what riders actually do, it is natural for cadence to vary. Riders rarely ride at a fixed cadence unless the conditions are also fixed (that is, same speed, same slope, same wind, same power).
    – R. Chung
    Commented Jan 28, 2016 at 6:37
  • 2
    Vote to leave open - OP is asking about some kind of sawtooth cadence pattern where rider pedals up to speed and then coast till you're too slow, then repeats. He's not asking about optimal cadences.
    – Criggie
    Commented Jan 28, 2016 at 10:27
  • 4
    Human muscle efficiency is less if you pedal intermittently vs pedaling continuously at a reasonably constant pace. And, given that wind resistance is proportional to the square of relative velocity, riding fast then slow is less efficient in that regard as well. Commented Jan 28, 2016 at 13:30
  • 1
    "...more efficient than if you accelerated slowly and maintained the 50 mph" if this is true it's more likely to be because your average speed is lower and if you sat steadily at the same average your fuel consumption would be lower again.
    – Holloway
    Commented Jan 28, 2016 at 14:24
  • 2
    @David efficiency should be more important to the older riders, getting the most efficient use from their muscles. And accelerating then coasting then accelerating over and over would not lead to a leisurely ride. You seem to be confusing efficiency, power and speed here Commented Jan 28, 2016 at 16:29

4 Answers 4


This is really a pretty complex question. Without knowing where you read about "steady cadence being a good thing" or what the author meant it is hard to evaluate this statement, but this SE.bicycles answer presents data showing that riders don't ride at a steady cadence. Rather, they alter their cadence according to conditions of the ride, the level of power they wish to put out, and the choice of gearing they have available. So your initial premise may not be correct.

That said, assuming you are on a constant gradient, with consistent wind, trying to go as fast as possible over a fixed distance, then you want power to be constant, because constant power application is the fastest way from point A to B in those conditions. If power is constant, then cadence will also be constant, unless you are shifting. Most humans can maintain the same power over a fairly wide variety of cadence, even if it doesn't feel as good. Usually it is fine to self select what cadence you want. It may be the case that you prefer a different cadence when climbing vs on the flats. If that is so, do not feel that you have to shift to keep cadence the same as the terrain changes.

The second part of your question presents a scenario of alternate pedaling and coasting, and posits that since it may be more economical for automobiles to do so perhaps it will be so for human powered cycling as well. However, humans are not like internal combustion engines -- we have two different metabolic pathways to generate power, which are colloquially called "aerobic" and "anaerobic." Aerobic metabolic production of ATP via the Krebs cycle is relatively efficient, but when we need to produce power that exceeds our aerobic capacity we switch to "anaerobic" production that is sustainable for far shorter periods of time. Compounding this "supply of power" issue is the "demand for power". The demand equations for cycling power are well understood if not well known, but the aerodynamic component of drag varies with the square of airspeed so the power demanded varies roughly with the cube of speed in calm conditions. These two physical constraints mean both that cycling faster and then coasting requires more energy than cycling steadily at the average pace, and also that speeding up is more physiologically taxing. That is, 1 minute of high power followed by 1 minute of recovery is still more taxing than 2 minutes at the average power, even though the total joules of work performed is the same. This is the basis for why "high-intensity" interval workouts lasting X minutes can have a greater training effect but be much less enjoyable than the same X minutes steadily spent at the average power.

As it happens, here is a plot based on data from a segment of a ride where the rider alternately pedaled and then coasted. This segment of the ride was done on a flat road, with almost no wind. The upper panel shows cadence; the middle panel shows speed; and the bottom panel shows power.

alternate pedaling and coasting

The "sawtooth" in cadence and power are apparent, and you can see speed increase and then slowly decrease. For this roughly 4 km long segment using this particular pattern of pedaling and coasting, power averaged 100 watts. If, on the other hand, the rider had covered this same flat 4 km long segment at constant speed so that the distance and total elapsed time were the same, the power would only have needed to have been 70 watts, a savings of 30%.

In this case, the bicycle used was a geared bike, so had the rider changed gearing he could have produced the required 70 watts with many different combinations of cadence and pedal force. In other words, steady cadence need not have been used. This is an illustration that cadence isn't terribly consequential, and you have wide latitude to ride at whatever cadence you want to meet the purposes you wish to achieve. If you enjoy pedaling and coasting, there nothing wrong with that.

  • 1
    Thanks Robert, I will just let your comment stand as a nice expansion.
    – jackmott
    Commented Jan 28, 2016 at 20:45
  • To me it is strange why humans would benefit from a constant cadence when we know cars are more efficient when they accelerate briskly (using about 75% of their maximum power) and then coast down (ideally with engine off but for safety lets say just with no throttle). This technique alone is worth about a 10% boost in fuel efficiency (not that it matters now with these low gas prices). If someone on a bike is not in a big hurry, the burst then coast technique might get them there slower but with far fewer crank rotations. I will experiment when I ride more and decide which I like better.
    – David
    Commented Jan 28, 2016 at 21:13
  • 3
    Unfortunately there is a nonlinear relationship between power output and physiological fatigue in humans, so constant power application is more efficient. This has already been tested a few hundred thousand times per year by various time trial experts, but grab a power meter and experiment, it is fun. That potential benefit of burst/coast realized with cars (sometimes) depends on subtle interactions with airflow that don't occur in humans.
    – jackmott
    Commented Jan 28, 2016 at 21:22
  • @jackmott interesting, what are the subtle interactions that occur in cars?
    – ebrohman
    Commented Jan 29, 2016 at 0:55
  • I believe one thing that you get from the pump and coast method is lower pumping losses, because when the engine is on power, the throttle body is fully open. This technique may not even work though on newer engine designs with better variable valve timing and/or no throttle at all. Easy to fool yourself into thinking it works because testing for the differences is hard. But that is straying way off topic!
    – jackmott
    Commented Jan 29, 2016 at 1:19

Another theory that does not seem to be adopted by the bicycle community.

Individual racers and race teams go with a steady cadence and output.

If you have two riders trying dump a third the 2 riders will take turns bursting and falling back. Eventually the 3rd rider cannot keep up with the burst and is left. If bursting was as (or more) efficient than a steady pace this would not work. This works even with the 3rd rider not taking lead - they are drafting.

Wind resistance is relative to velocity squared.
Split 10 mph and 30 mph versus 20 MPH
2*20*20/(10*10 + 30*30) = .80
That is right 20% less wind resistance

As for muscle fatigue. Fatigue is the force - not cadence. Drop a gear so it is easier to spin. It is well establish that a human body is more efficient at delivering power in a fairly narrow cadence range.

  • Somehow I am not "buying" this steady is better thing. I work a 8 hour shift for example. During that shift I lift some heavy things repeatedly which tires me some. If you were to ask me what would make me more tired, working hard (fast paced) and taking a break every 2 hours or slightly slacking (working a little slower) but maintaining that speed for the entire 7.5 hours, the no break would easily tire me way more! Even going 1 hour past my scheduled break (working 3 hours straight) is hard on me. Also batteries exhibit this same properly. Take a load off them and they recover quickly.
    – David
    Commented Jan 28, 2016 at 21:19
  • Maybe the actual amount of energy expended is less if a biker maintains constant cadence but maybe the perceived fatigue without resting is that MORE energy is expended. Why would it matter if less actual amount of energy was expended if it feels like more was? When I get to my destination I would like to be able to walk like a normal person. I suspect the majority of older riders need this rest, especially with old tired bones. Why do you think they take naps midday? You expect a 50+ year old man to maintain an 80 cadence for a 6.5 mile (10km) ride? If I had a NuVinci hub I could check.
    – David
    Commented Jan 28, 2016 at 21:26
  • 1
    @david Then don't buy it. Once again you have a strategy we don't see used in the industry. Have you ever seen a team take a break in the Toure de France or any race? Now this a 50+ year old man. I am 50+ year old man and 6.5 miles is walk in the park - 16 is a standard workout and I don't take a break. I said narrow an older person my be a little lower than 80 but there will be a narrow optimal range. I did not post this answer for you. I am not going to answer any more of your comments.
    – paparazzo
    Commented Jan 28, 2016 at 21:29
  • 1
    Squared is squared. Wind resistance should be less - did you run the calc? Yes we are talking about radially different situations. I am talking about reality and you are not. Your sense of logic and analytical reasoning is a bit lacking to be polite.
    – paparazzo
    Commented Jan 28, 2016 at 22:33
  • 1
    I cannot understand how you would think a leisurely ride at a steady 10mph with legs spinning at a nice easy cadence would be less leisurely than racing up to 20mph, getting the cardio vascular system and muscles working hard then coasting down to an almost standstill to repeating till the end of the ride. This is interval training and is anything but leisurely Commented Jan 29, 2016 at 6:57

David - how does it feel to you when you're on your bike and try doing exactly that?

Biomechanically, you'll be stressing your muscles to a higher peak level, and then a brief respite. I imagine your overall power output will be lower over a reasonable distance.

If you pedal fast then coast for a bit then pedal then coast - I think you'll look like a teenager on a kid's BMX, and your average speed will be pretty slow. Partially this is because they're seriously under-geared for their teenage power, and partially its a lack of endurance to maintain those cadences for longer periods.

Answer Get on your bike and ride!

  • 1
    I'm not sure if this actually answers the question at hand. The first paragraph is relevant, but in order to answer the question, it would also be necessary to explain why stressing your muscles to a peak level and then resting is bad. This actually sounds a lot like intervals, which are widely accepted as a great training practice. The second paragraph basically boils down to "it looks silly," which is just an opinion (though not one I disagree with), not a reason that steady cadence is important.
    – jimchristie
    Commented Jan 28, 2016 at 13:22
  • 1
    @criggie I somehow seem to have downvoted you on my stupid phone, I had up voted this previously. I can't change until you edit it though. Commented Jan 28, 2016 at 18:00
  • 1
    @CearonO'Flynn Its not about the votes, its about finding the answer.
    – Criggie
    Commented Jan 28, 2016 at 20:16
  • 1
    @Criggie I know but I thought it was a good answer and didn't want to do this answer a disservice Commented Jan 28, 2016 at 20:25

It is far more efficient to cycle at a steady cadence suitable for the terrain and gearing used. In fact this attempt to maintain cadence and maximise efficiency is the reason bikes have gears.

Varying the cadence is going to put stress and strain on your muscles, joints and cardio vascular system. This could be your aim (with the exception of stressing your joints) as this is what you would do with interval training, however it would not be usual for a normal ride where you want to get from A to B the most efficiently.

I would also imagine gear changes would be be carried out with strange loads on them, either too much when cadence is high, or too low when the cadence is down. This could stress the mechanical components of the bike too.

  • I feel good when I coast and get a rest. I even like to pedal (freewheel) backwards occasionally cuz it helps my muscles. To me it is not about maximizing efficiency, it is more about what feels good and if I get a burst of energy to use it and if I get tired to coast for a while to recover. I ride my bike for the enjoyment of it (and also for practical transportation), not to try to set any speed or efficiency records.
    – David
    Commented Jan 28, 2016 at 16:26
  • I find I enjoy my ride far more if I have a steady pedal stroke, my muscles and joints don't hurt and it means I can ride further and for many more days in a row. If I want to punish myself and have a hard muscle breaking training ride that I'd feel the next day I would pedal as you describe, varying cadence, accelerating hard then resting and repeating over and over Commented Jan 28, 2016 at 16:32
  • Single speed bikes don't easily allow for steady cadence since there are varying loads (uphill, downhill, paved, dirt, headwind, tailwind...). So if that is the case, why would varying cadence on a multispeed bike be bad? Our muscles change load all day long. We stand, sit, lie down, walk, sometimes run... I think it is fun to ramp up to speed then give myself a rest. Pedaling the same cadence the entire time would be kinda boring unless maybe I had a NuVinci hub and locked it at my favorite cadence of 88 (at which feet per crank rotation and MPH match exactly).
    – David
    Commented Jan 28, 2016 at 16:40
  • 1
    @david a geared bike has gears for this very reason, to keep the cadence the same as much as possible, because it is easier, more efficient and gentler on joints & muscles. They are two different machines and two totally different riding experiences can be achieved with them. The answer to your question is because its more efficient you may not agree with it but that is the reason, and its why gears were invented Commented Jan 28, 2016 at 16:45
  • 4
    @David You say in your comment: "To me it is not about maximising efficiency" but in the question, you say " I would think it would be more efficient to accelerate briskly then coast down, then repeat." Are you asking about efficiency? Are you really asking anything?
    – brendan
    Commented Jan 28, 2016 at 22:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.