Glance at your bike computer long enough and you run into a strange contradiction. The cadence that tests as most efficient in a laboratory is not the cadence trained cyclists actually ride. Lab work keeps landing on roughly 60 rpm as the metabolic sweet spot, yet fit riders settle in around 90 rpm on their own, and elite pros spin faster than that. This guide sorts out why. You'll get the newest research (including a 2024 study that finally pinned down how your ideal cadence shifts with effort), a clear target for whatever situation you're in, and drills to widen the range where you feel smooth.

Key takeaways

- There is no single "best" cadence. The most metabolically efficient rate is low (~60 rpm), but trained riders self-select ~80–100 rpm because efficiency is not the only thing that matters.

- Your optimal cadence rises with intensity — from roughly 66 rpm at an easy aerobic pace to about 84 rpm at VO2max, per a 2024 study.

- Power = Torque × Cadence. At a fixed power, low cadence means high force per stroke (muscular and joint load); high cadence means low force but a bigger cardiovascular cost.

- Deliberate low-cadence intervals can build fitness fast: one 2024 trial saw nearly double the VO2max gain versus free-cadence training.

- Aim for a range, not a number — and use targeted drills to expand both ends of it.

The 60-vs-90 cadence paradox

Cyclists have argued about cadence for decades, and the argument keeps going because both camps are right about different things. Put riders on an ergometer and measure the oxygen cost per watt, and the cheapest number, the true metabolic minimum, usually sits down around 60 rpm. Depending on the workload it can land anywhere in the 50–70 rpm window. That finding goes back to Coast and Welch's 1985 work and has been replicated plenty of times since. If fuel economy were all that mattered, we'd all be grinding along in a big gear.

But nobody actually races that way. Trained cyclists pick something much brisker, roughly 80–100 rpm, with the elite crowd hovering near 90 rpm on hard efforts (Lucía and colleagues measured exactly this in Grand Tour pros back in 2001). There's the paradox in a sentence: the cheapest cadence and the chosen cadence sit about 30 rpm apart, and the riders choosing the faster one are the fittest people on the planet.

The way out of the contradiction is that oxygen economy is only one thing your body is trying to solve for. Studies that pull the goals apart find two different optimums pulling in opposite directions. Bieuzen and colleagues, for instance, measured a most-energetically-optimal cadence of about 64.5 rpm, but a neuromuscular optimum, the cadence that minimizes muscle force and strain per stroke, of about 93.5 rpm in well-trained riders. Your legs and your lungs want different things, and the cadence you fall into naturally is the truce they've negotiated.

So here's the practical upshot. "Most efficient" is a lab answer to a lab question. Out on the road you're juggling muscle fatigue, joint load, heart rate, and how long you still have to keep going, all at the same time. By the end of this guide you'll know which of those to protect in each situation, and how to nudge your natural cadence toward the range that serves you.

What cadence actually is, and the one equation that explains everything

Cadence is just how many full pedal revolutions you complete in a minute, measured in rpm. That's the easy part. The part worth remembering is the equation that drives everything downstream:

Power = Torque × Cadence

Power is what moves you down the road, and it's the product of how hard you push each stroke (torque) and how fast you turn the cranks (cadence, as angular velocity). Because it's a product, there are endless ways to make the same number. You can put out 250 watts by mashing a big gear slowly with a lot of force, or by spinning a smaller gear quickly with very little force. Same watts, completely different demands on your body.

That's the trade-off at the heart of the whole topic, and it isn't a matter of taste. It's arithmetic:

• Lower cadence means higher torque per stroke. Each push takes more force, which loads your muscles and knees harder but keeps your heart rate down.
• Higher cadence means lower torque per stroke. Each push is gentler on the muscles and joints, but the faster leg turnover drives your heart rate and breathing up, moving the cost onto your cardiovascular system.

Neither one is "correct." What you're really choosing is where to spend the effort, in your muscles or in your lungs. A rider with a big engine but cranky knees should lean toward spinning. A rider who's gasping on a long climb might grind for a bit to keep heart rate in check.

Key takeaway: Every other decision in this article, climbing, sprinting, injury, drills, is just this one equation applied to a specific case. Once it clicks that cadence redistributes load rather than handing you free speed, most of the "right number" argument dissolves.

What the science calls "most efficient," and why it isn't the answer

Read only the abstract of the classic cadence studies and you'd walk away thinking the ideal is around 60 rpm, full stop. The oxygen-cost curve really does bottom out low. So why does every experienced coach tell you to spin faster than that?

Because your freely chosen cadence (FCC) minimizes perceived exertion, not oxygen consumption, and those are two different targets. A well-regarded review of cadence and neuromuscular function describes it as a balancing act along a U-shaped curve. Pedal below your freely chosen cadence and you save a little oxygen but pay for it in higher mechanical joint torque and muscular strain. Pedal above it and you cut the force per stroke but the oxygen cost climbs. Your body parks itself at the bottom of the combined discomfort curve, and that spot sits well above the pure metabolic minimum.

There's a fuel-and-fibre story hiding behind the oxygen numbers, too. At a hard 85% of VO2max, riding at 50 rpm burned through more fast-twitch glycogen and produced higher blood lactate than riding at 100 rpm, precisely because those low-cadence, high-force strokes lean on the powerful, glycogen-hungry fast-twitch fibres. So a cadence that's "efficient" at the whole-body oxygen level can quietly be expensive at the level of the exact muscle fibres you'll want later in the ride.

This is why the honest answer to "what's the best cadence for cycling?" is that it depends on what you're optimizing and how hard you're going. There really are two optimums, a metabolic one near 60–64 rpm and a neuromuscular one near 93 rpm, and your self-selected cadence is your body splitting the difference in real time. Chase the lab's 60 rpm on every ride and you'll end up with heavy, prematurely tired legs even while your oxygen meter looks tidy.

Decision rule: Don't try to force your cadence down toward the metabolic minimum for normal riding. Treat your comfortable self-selected cadence as the baseline, then move it up or down on purpose for the situation. The next section puts numbers on exactly that.

What's new in 2026: your optimal cadence moves with intensity

The biggest recent shift in cadence science is really a shift in the question. It's moved on from "what's the one best number" to "your best number changes with how hard you're riding." A 2024 study in Frontiers in Physiology (and the Journal of Science and Cycling) finally attached precise figures to the idea. Optimal cadence rises sigmoidally with intensity, stepping up through each physiological threshold:

Table 1 — Optimal cadence by exercise intensity (2024 Frontiers in Physiology study)

The differences held up statistically (p < 0.01). Here's the part that surprised me: sprinters and endurance athletes showed no difference in optimal cadence at each threshold, apart from that fatigue-free maximum. The curve is basically universal. Your discipline or your muscle-fibre makeup doesn't hand you a private ideal cadence at a given intensity.

Why does the ideal rate climb as you push harder? It comes down to the successive recruitment of faster-twitch fibres, which have higher optimal contraction velocities. Building on Henneman's size principle and the work of Sargeant and Zoladz, the picture is simple enough: harder effort recruits faster fibres, and faster fibres pedal best at faster cadences. Your legs literally re-tune their preferred speed as you add watts.

Two more recent findings fill in the 2026 picture. A 2024 PLOS ONE trial (31 trained cyclists, 8 weeks, matched interval sessions) found that a low-cadence group training at 40–60 rpm improved VO2max by +8.7% versus +4.6% for the freely-chosen-cadence group, and maximum aerobic power by +8.1% versus +3%, close to double the adaptation per session. Then a 2025 study showed that pushing cadence above a standard 60 rpm did not change VO2max or aerobic ceiling, but it did raise ventilatory demand and heart rate while lowering maximum work rate. The thread running through all three: cadence changes which system you stress, not your ceiling fitness. That's what makes it a training lever, not just a comfort setting.

High vs low cadence: when each one wins

Turn all of that science into a field guide and you get a clean situational framework. The rule of thumb below is backed by TrainerRoad's coaching data and lines up with the intensity curve above:

Table 2 — Situational cadence targets

Treat this as a decision framework, not a rulebook. Four questions get you to the right number:

1. How long does this effort have to last? The longer it runs, the more you want to protect your muscles by keeping cadence in the moderate 70–90 band and letting your cardiovascular system carry more of the load over time.
2. How hard is the effort right now? Match the intensity curve. Easy endurance sits comfortably in the 70s, threshold work in the low 80s, all-out efforts higher.
3. What's your limiter today? Cramping or heavy legs, spin faster to unload the muscles. Gasping on fresh legs, briefly drop cadence to ease the cardiovascular cost.
4. What are you training? If the goal is a training stimulus rather than comfort, break these ranges on purpose (more on that below).

That last question is the important one, because grinding isn't a mistake. It's a tool. The 2024 PLOS ONE result means deliberate low-cadence work at a moderately hard effort forces fast-twitch fibres to build oxidative capacity, an adaptation that's otherwise hard to reach on a bike. So the "high vs low" question has two answers. For riding well today, follow the situational table. For building specific fitness, program low-cadence intervals on purpose.

What cadence do the pros actually use?

The pro peloton is where cadence turns into personality, and the classic case study is the early-2000s duel between two Tour de France rivals. Lance Armstrong was the archetypal spinner. His coach Chris Carmichael had him climb L'Alpe d'Huez in a 39×23 or 39×21 at around 90 rpm, and reporting from the era put him at 95–110 rpm on long climbs while rivals slogged along near 70. Jan Ullrich got cast as the grinder, famously described mashing something like 65–75 rpm on the same climbs. Worth an honest caveat here: some measurements suggest Ullrich was actually closer to 90 rpm and the contrast got played up for television. The legend is cleaner than the data usually is.

Chris Froome carried the high-cadence style into the next era, with a well-documented climbing rhythm in the 90–100+ rpm range, using low torque per stroke to protect his legs across three weeks of racing. Modern WorldTour norms have crept upward, and the discipline-specific bands shake out like this:

Table 3 — Cadence by rider and discipline

Elite sprint cadences of 120–140 rpm have been reported in the lab (Abbiss; Ansley and Cangley), which fits those track figures. One more honesty note, the kind that separates a trustworthy guide from a padded one: per-stage cadence for today's biggest names, Tadej Pogačar and Jonas Vingegaard, isn't publicly documented. The media reports their power and watts-per-kilogram, not their rpm. Their styles put them broadly in the ~85–100 rpm climbing band, but anyone quoting an exact stage cadence for them is guessing.

Key takeaway: The pros don't share one magic cadence. What they share is a bias toward spinning at whatever intensity they're riding, because protecting the muscles over a long race beats squeezing out a sliver of lab efficiency.

Does low cadence wreck your knees? The biomechanics

This is the section most cadence articles skip, and it's where Power = Torque × Cadence stops being abstract. For a given power output, lower cadence means higher crank torque, which means a higher knee-extension moment, which means a higher patellofemoral joint reaction force. In plain English: the harder you push each stroke, the more compressive load you drive through the front of your knee. Low cadence doesn't "damage" a healthy knee on its own, but it clearly increases joint loading, and that matters if you're already sore, older, or ramping up volume in a hurry.

The clinical advice follows straight from that. Cycling Weekly's expert coverage warns that jumping straight into very low-cadence work, 40–50 rpm, can bring on pain at the front of the knee (patellofemoral pain). The safer route is to start low-cadence drills around 70 rpm and only work lower as your tendons and muscles adapt. There's an age dimension as well. Riders in their 40s, 50s and beyond should grind less, because declining bone mineral density and joint resilience make high-torque strokes a worse bet than they were at 25.

If you're already dealing with anterior (front-of-knee) pain, the fix is basically the problem in reverse: shift to a higher cadence of 85–95 rpm in an easier gear. You keep making the power you need but cut the force per stroke, which unloads the joint. The evidence here leans on solid biomechanics plus clinical experience rather than large prospective injury trials, which are rare in cycling, but the mechanism is direct and the fix costs nothing to try.

Knee-protection checklist:

• Never start low-cadence intervals cold. Begin any high-torque work above 70 rpm and drop the rpm gradually across weeks, not within a single session.
• If your knees hurt, spin, don't mash. Drop to an easier gear and lift cadence to 85–95 rpm for the rest of the ride.
• Scale torque to your age and history. Masters riders and anyone returning from injury should favour cadence over gearing on climbs.
• Watch for the warning sign. Sharp pain at the front of the kneecap under load is your cue to shift up a gear and add rpm right away.

How to find and measure your cadence in 2026

Before you can improve your cadence, you have to know it. Start by finding your honest self-selected baseline, which tends to track your fitness. Rough bands by rider level: beginners often sit as low as 50–60 rpm and up toward 85; intermediate riders land near 60–80 rpm; experienced riders and racers cruise at 80–100 rpm; and pros push past 100 rpm in attacks and top 110 in sprints. None of that is a value judgment. A 2025 study in the Journal of Sports Sciences confirmed that elite cyclists pedalled significantly faster than novices across every measured interval (p < 0.001). Your comfortable cadence drifts upward as your training status improves, so a rising natural cadence is a quiet sign you're getting fitter.

You've got two ways to measure it.

The free way is manual counting, no sensor needed. Count one leg's downstrokes for a fixed window and multiply: 15 seconds × 4, or 10 seconds × 6, or 30 seconds × 2, they all give you rpm. It's clumsy mid-ride, but it's perfect for a quick baseline check on the trainer.

The precise way is a sensor or a power meter. For live, continuous data you want a dedicated cadence sensor or a power meter (which measures cadence natively). Here's the 2026 gear reality:

Table 4 — Cadence measurement options and 2026 pricing

Decision rule for gear: If you already own a power meter, you don't need a separate cadence sensor. It's already tracking rpm. If you don't own one, an $18 Magene S314 is all the hardware most riders will ever need to train cadence deliberately. Spend the four figures on a power meter only when you want the full power-training picture, not just cadence.

Key takeaway: Establish your baseline first, even by counting, because every drill below is prescribed relative to your current cadence, not to some absolute number.

Cadence drills to widen your efficient range

The point of cadence training isn't to bully yourself up to 100 rpm overnight. It's to widen the range where you feel smooth and efficient, so you can spin fast when you attack and grind strong when you climb. Here are four evidence-based prescriptions, each aimed at a different adaptation.

1. Spin-ups / fast-pedal drills (build leg speed). In a small gear (around 39×15), spin from a rolling start up to your maximum non-bouncing cadence and hold it for 20–30 seconds, then recover 3–5 minutes. Repeat 6–10 times. The cue that makes these work: keep your upper body relaxed and don't bounce in the saddle. If you bounce, you've hit your neuromuscular limit and should back off a touch.

2. High-cadence tempo intervals (endurance leg speed). Ride 3×15 minutes at 100–110 rpm (or roughly 10–15 rpm above your normal), staying in a small gear and capping the effort at Zone 3. Recover 5 minutes easy between reps, and do 1–2 sessions a week. This teaches your cardiovascular system to handle fast turnover without redlining.

3. Low-cadence muscle-tension intervals (strength and fibre recruitment). In a big gear at 50–55 rpm, hold about 85–95% of FTP (below threshold), building from 4×5 minutes for beginners (~20 min total) up to 4×10 minutes for advanced riders (~40 min total), with recovery equal to the interval length. The research-backed variant from the 2024 PLOS ONE study is 4×4 minutes at 40–60 rpm, RPE 7, on a 4–7% climb, with 4-minute recoveries, one session a week for a 4–8 week block. That protocol delivered nearly double the VO2max gain per session by forcing fast-twitch fibres to build oxidative capacity.

4. The beginner ramp (raise your baseline safely). If your natural cadence is low and you want to lift it, nudge it up by just +3–5 rpm and hold for 5 minutes. If your heart rate climbs more than a few beats, back off and try again later. Small progressive steps beat white-knuckling 100 rpm and quitting after a minute.

Across all four, use the "kick and pull" cue for a smoother stroke: near the top of the pedal circle, lightly kick your toes into the fronts of your shoes; near the bottom, pull your heels back into the heels of your shoes. It smooths out the dead spots and makes a high cadence feel less frantic.

Drill-selection framework: Want raw top-end speed for sprints? Prioritize spin-ups. Want to raise your everyday cruising cadence? Use the beginner ramp plus high-cadence tempo. Want more climbing power and a VO2max bump? Program the low-cadence muscle-tension block, carefully, per the knee guidance above.

Frequently asked questions

Q: Is 90 rpm too high for me? A: Almost certainly not. Around 90 rpm is what most trained cyclists naturally choose and what the neuromuscular optimum (~93 rpm) points to, because it minimizes muscle force per stroke. If 90 feels frantic, that usually reflects fitness and habit rather than a wrong target. Use the beginner ramp to close the gap +3–5 rpm at a time.

Q: Why do I naturally pedal slowly, around 60–70 rpm? A: That's completely normal, especially for beginners, whose self-selected cadence often sits in the 50–70 rpm range. Lower cadence loads the muscles more and the cardiovascular system less, which can feel easier while your aerobic fitness is still developing. As you get fitter your comfortable cadence tends to drift upward on its own.

Q: Does a higher cadence make you faster? A: Not by itself. Because Power = Torque × Cadence, spinning faster in an easier gear at the same force just gives you the same (or less) power with a higher heart rate. Cadence is a tool for redistributing load between muscles and lungs, not a free speed boost. Speed still comes from producing more power.

Q: What cadence should I use for climbing? A: Most riders and pros climb best by spinning in the 85–100 rpm range in an easier gear, which protects the legs by keeping per-stroke force down. Grinding a big gear at 60 rpm raises knee load and burns fast-twitch glycogen faster. Deliberate low-cadence climbing is valuable as training, but for everyday climbing, spin.

Q: Does cadence still matter if I train with a power meter? A: Yes, and your power meter already measures it. Power tells you the output; cadence tells you how you're producing it. Two rides at the same watts can stress your muscles or your cardiovascular system very differently depending on cadence, so it stays a lever worth watching and training even when power is your primary metric.

Q: Is low-cadence grinding bad for my knees? A: It increases knee joint loading, because higher torque per stroke raises the patellofemoral joint force. A healthy knee handles this in moderation, but if you have front-of-knee pain, shift to 85–95 rpm in an easier gear. Never launch into 40–50 rpm work cold. Start low-cadence drills around 70 rpm and progress gradually, and grind less as you get older.

Q: Do low-cadence intervals really build more fitness? A: The 2024 PLOS ONE trial suggests they can. A 40–60 rpm group gained +8.7% VO2max versus +4.6% for free-cadence training over 8 weeks, nearly double per session, by driving fast-twitch fibres to develop oxidative capacity. They're a potent, time-efficient stimulus, but they demand respect for the knee-load caution above.

The bottom line

Optimal cycling cadence isn't a number you can print on a sticker. It's a range you manage. The lab's ~60 rpm minimizes oxygen cost, the neuromuscular optimum near ~90 rpm minimizes muscle strain, and the 2024 research shows your best rate sliding from about 66 rpm at easy efforts to 84 rpm at your aerobic ceiling. Trained riders and pros bias toward spinning because protecting the muscles beats chasing lab efficiency, and the same physics (Power = Torque × Cadence) explains why grinding loads your knees while spinning taxes your lungs. Find your honest baseline, respect your joints, use the situational targets to ride smarter today, and program the drills to widen your efficient range. That, and not some single magic rpm, is how you spin faster and stronger in 2026.