Two riders roll side by side at identical speed, one turning the pedals at 60 rpm in a big gear, the other at 95 in a smaller one. Externally they are doing the same thing. Internally they are doing different sports: the grinder is doing repeated leg presses, the spinner is doing cardio. This article explains what cadence really trades off, why the 80–95 rpm band became the consensus for most riding, and the situations where low cadence is the right tool rather than a bad habit.
Cadence has a learning curve worth warning about: the first sessions of deliberate spinning feel inefficient and faintly ridiculous, the legs bouncing, the saddle complaining. That is a coordination problem, not evidence. It resolves in weeks, and the resolution is permanent — which is more than can be said for most cycling improvements.
Force versus frequency
Power is force times cadence. Produce a given power at 60 rpm and each pedal stroke must push half again as hard as at 90 rpm. Muscles fatigue with force: high-force strokes recruit fast-twitch fibres, drain glycogen locally and load knees and tendons with every revolution. Shift the same power to a higher cadence and the load moves to the heart and lungs, which are built for continuous work and recover in minutes rather than days. That is the whole argument for spinning in one paragraph: it moves the effort from the system that wears out to the system that adapts.
Key takeaways
- Power = pedal force × cadence; you choose the mix
- High force per stroke fatigues muscle and loads joints
- High cadence shifts cost to the cardiovascular system
The consensus band, and its evidence
Observational studies of trained cyclists consistently find freely chosen cadences around 85–95 rpm on flat ground, and laboratory work shows gross efficiency is often slightly better at lower cadences — a paradox resolved by distinguishing efficiency from durability. The lab measures the first minutes; the road lasts hours, and riders unconsciously protect their muscles for the duration. Beginners typically self-select 55–70 rpm because spinning feels unnatural before it feels easy — coordination, not fitness, is the barrier for the first weeks. A cadence sensor is a genuinely useful teacher here, and our gear calculator shows what each cadence does to your speed in each gear while you learn.
When grinding is right
Three honest exceptions. Standing climbs: out of the saddle, body weight does the work and cadence in the 60s is normal and sustainable in short doses. Technical mountain biking: a half-stroke at low cadence ratchets over rocks and roots where spinning would strike pedals. And deliberate strength work: coaches prescribe short low-cadence intervals as on-bike resistance training. What these share is brevity and intent. Grinding as a default, all ride, every ride, is not strength training; it is just how knees get sore.
Cadence and the saddle: bounce is data
The tell-tale of cadence beyond your current skill is bouncing in the saddle — your legs are pushing the pedals faster than your coordination can smooth, and the excess becomes vertical motion. The fix is not to abandon the cadence but to approach it in steps: raise your cruising cadence five rpm at a time, letting each step feel normal before the next, and practise smoothness drills — gentle high-cadence minutes on safe, flat roads — rather than living at the edge of control. Pedalling in circles is a coaching fiction, but pedalling without dead spots is a trainable reality, and it is what high cadence actually consists of.
Gear choice enables all of it: you cannot practise 95 rpm if your gearing at cruising speed only offers 78. That is a specification problem, and the gear calculator settles it before the ride does.
Sprinting and standing: cadence’s loud relatives
At the extremes, cadence tells the whole story of a ride. A sprint finish is cadence weaponised — track sprinters exceed 130 rpm because at maximum speed there is no bigger gear available in the moment, only faster legs. Standing climbs are the opposite pole: the rider leaves the saddle, lets body weight drive the downstroke, and cadence drops into the 60s by design. Both are correct, and both are short. The error is importing either into the middle distance of a ride — sprint cadence as a cruising habit, or standing-grind as the response to every rise. The middle is where tours are completed and knees are preserved: seated, spinning, unremarkable, and faster at the end of four hours than any heroics the first hour offered.
If you remember one line: force is spent, cadence is rented — and you get the rent back at the top of every hill.
Cadence for new riders: a four-week on-ramp
If your natural cadence today is in the low 60s, jumping to 90 by willpower alone produces bounce, frustration and a quiet return to grinding by week two. The on-ramp that works: week one, simply measure — a cheap sensor or fifteen-second counts — and ride as you ride, noting the number on flats and climbs. Week two, ride one gear lighter than instinct on the flat and let cadence rise by whatever it rises. Week three, choose two ten-minute blocks per ride to hold cadence five beats above your week-one average. Week four, repeat with another five. The month costs nothing but attention, and the endpoint — a default in the low 80s with climbs in the 70s — is where most riders’ knees send their first thank-you note.
Revisit the gear calculator at the end of the month: the same gears now produce meaningfully different speeds in your hands, which is the cleanest demonstration available that you, not the bicycle, were the variable.
Building the habit
Change one gear lighter than instinct for a fortnight and let your legs recalibrate; the first rides will feel absurdly twiddly and slightly slower, and both sensations are temporary. Count pedal strokes for fifteen seconds at intervals (×4 = rpm) if you have no sensor. On climbs, change down early — before cadence craters — because a gear changed under full load at 45 rpm is a small mechanical protest you can hear.
Common mistakes
- Choosing gears by the speed you wish you were doing
- Standing to grind as the first response to every gradient
- Judging cadence advice by a 20-minute lab test instead of a four-hour ride
Questions this raises
No — above ~105 rpm most riders bounce and waste energy on coordination. There is a personal ceiling; find the band where you are smooth.
On the flat, commonly yes — and higher in sprints. Their climbing cadence drops into the 70s and 80s like everyone else’s, which tells you the band moves with terrain.
Put a number on it.
The same speed can cost your knees or your lungs. The physiology of pedalling speed, the 80–95 rpm consensus, and the climbs where grinding is honest work.
Open the Gear & Cadence Calculator