The Moveee team
Free coaching · a real route for every ride
The question gets asked defensively, usually by someone who already wants the answer to be no. It deserves a better response than either of the usual ones — that e-bikes are cheating, or that they are just as good as a real bike.
The honest answer is more interesting than either. Electrically assisted cycling is measurably moderate-intensity exercise for most people riding it, it does improve fitness in people who were previously inactive, and it makes almost no difference at all to a trained cyclist's fitness unless it changes how much they ride. Those three statements are all supported by the literature and they are not in conflict.
Here is what has actually been measured, what the trials found, and where an e-bike does and does not fit alongside training.
How hard is e-bike riding, really?
Assistance reduces the power you have to produce, but it does not reduce it to zero, for a simple reason: on a pedal-assist bike the motor only contributes while you pedal, and assistance is capped by law at a speed most people would otherwise ride below. So you keep pedalling, and you go faster, and the net effect is a smaller reduction in effort than people assume.
When this has been measured directly rather than guessed at, e-bike riding comes out as moderate-intensity activity — lower than conventional cycling on the same route, higher than walking, and above the threshold generally used to count activity as health-relevant. Comparisons of heart rate, oxygen uptake and perceived exertion between assisted and unassisted riding show the same ordering consistently.
Stylised. The ordering and the placement relative to the moderate threshold reflect the published comparisons; the bar heights are illustrative rather than pooled effect sizes.
That placement matters because of what the World Health Organization's physical activity guidelines ask for: at least 150 to 300 minutes of moderate-intensity activity a week for adults, or 75 to 150 minutes of vigorous. An e-bike commute of half an hour each way, three days a week, lands inside that target on its own. For a person doing nothing at all, that is a meaningful change.
What the trials found
Several small trials have put previously inactive people on e-bikes for a few weeks and measured what happened. The results are consistent in direction and modest in size.
Höchsmann et al., 2018
32 overweight, inactive adultsFour-week randomised trial: e-bike commuting versus conventional bike commuting, at least six kilometres a day, three days a week.
Both groups improved peak oxygen uptake. The e-bike group's improvement was not statistically different from the conventional bike group's.
Peterman et al., 2016
20 inactive commutersFour weeks of pedelec commuting, at least 40 minutes a day, three days a week, with no other changes asked for.
Improved cardiorespiratory fitness and improved two-hour glucose tolerance. Riding intensity sat in the moderate range for most of each commute.
Langford et al., 2017
Instrumented comparisonMeasured physical activity intensity of pedal-assist e-bikes against walking and against conventional bicycles in real use.
E-bike riding was of moderate intensity — lower than conventional cycling, but higher than walking, and above the threshold conventionally used for health-related activity.
Bourne et al., 2018
Systematic review, 17 studiesPooled the health evidence on electrically assisted cycling available at the time.
E-cycling reliably provides at least moderate-intensity activity and can improve cardiorespiratory fitness in inactive people. The review was explicit that the study quality was generally low and the trials small.
The Höchsmann result is the one worth sitting with. Over four weeks, in overweight and inactive adults, e-bike commuting improved fitness and the improvement was not distinguishable from that of conventional bike commuting. That is a small, short trial and it should not be over-read — but it does undercut the assumption that assistance removes the training effect.
The equally important caveat comes from the systematic review that pooled this literature: the studies are small, mostly short, and generally of low methodological quality. Nobody has run a large, long, well-controlled trial of e-cycling and health outcomes. The direction of the evidence is clear. The precision of it is not, and anyone quoting a specific percentage improvement is over-claiming.
The substitution question
Here is where the honest answer gets less comfortable. Everything above concerns people who were doing little or nothing. The picture for a trained cyclist is different, and the mechanism is straightforward.
Training adaptation depends on stimulus. A ride that sits below the intensity you already train at, on a bike that removes the hard parts, does not add anything to an already-trained aerobic system. For a rider doing eight hours a week at a mix of intensities, an assisted ride is a pleasant afternoon, not a training session. If it displaces a real endurance ride, it is a net negative.
The word displacement is doing the work in that sentence. An e-bike ride that happens instead of a training ride costs you. An e-bike ride that happens instead of driving, or instead of sitting down, adds. The bike does not determine which of those it is — your week does.
There is also a real question about whether e-bike owners ride their conventional bike less. The travel-behaviour research finds e-bike owners riding further and more often overall, which points the other way, but most of that work is about transport rather than sport, and it does not tell you what a club rider's training week looks like after they buy one.
Where an e-bike earns its place
The strongest case, and the one with the best evidence behind it. Assistance removes the reasons people stop bike-commuting — arriving sweaty, a hill on the way home, a 15 km distance that is fine once and grim five times a week. A scoping review of e-cycling and travel behaviour found e-bikes are associated with longer trips and more of them. Four moderate hours a week that actually happen beat two hard hours that do not.
The most common failure in amateur training is that easy days are not easy. An e-bike makes an easy day structurally easy — you can cap the effort with the motor instead of with willpower on a headwind. The catch is that you have to actually want low intensity, and the assistance can just as easily take you below any useful stimulus.
This is the use case that does not need justifying by physiology. A partner who rides at 18 km/h and one who rides at 30 km/h cannot ride together on equal bikes without one of them having a miserable time. An e-bike under the slower rider solves that completely, and the ride happens instead of not happening.
Assistance lets you control load precisely at the low end and bail out of a headwind or a climb that your knee is not ready for. It is time on the bike with a ceiling on the effort, and it keeps the habit alive during the period when the habit is most fragile.
Cargo and utility e-bikes turn trips that were never going to be cycled into trips that are. Not training, and not pretending to be — but it is activity in a life that otherwise had none in it, which is where the health benefit of cycling mostly comes from.
Riding a course to learn its corners and gradients does not require you to be in bits at the top. Assistance lets you look at the road instead of at your stem.
Two of those deserve expanding. The commuting case is the strongest because it is the one where the alternative is usually zero — and we have made the general argument for commuting as training separately, in commuting by bike. The difference an e-bike makes is to the consistency, not to the physiology: a 14 km commute with a hill is something most people do twice and then stop doing.
The return-from-injury case is the other one. Getting back on the bike after a layoff fails most often because the first weeks are demoralising, and assistance takes the demoralising part out — you can ride the route you used to ride, at the pace you used to ride it, while your actual fitness catches up. Getting back on the bike after a break covers the rest of that process.
The limits, stated plainly
- It will not raise a trained cyclist's FTP. Nothing in the literature suggests otherwise, and the mechanism makes it implausible. Do not expect it to.
- It is a poor substitute for structured intensity. If your week needs a threshold session, an assisted ride is not it. Controlled intervals are easier to do on a trainer than on any road bike, assisted or not — zone 2 on a smart trainer makes the case for the low end, and the same logic applies upwards.
- "Easy" can become "nothing". An e-bike recovery ride only works if you are still pedalling meaningfully. Assistance set high on flat roads produces an hour of sitting outdoors, which is pleasant and is not a recovery ride.
- The evidence base is thin. Small trials, short durations, mostly inactive participants, low study quality by the reviewers' own assessment. The direction is consistent; the magnitude is not established.
- Weight, range and cost are real constraints. A 25 kg bike with a flat battery is a considerably worse bike than a 10 kg one, and the day you find that out is usually the day you needed it most.
The sensible conclusion
If you currently do not exercise, an e-bike is one of the better interventions available, because it produces moderate-intensity activity in a form that people keep doing. The trials in inactive adults show fitness improving over a few weeks, and the reason is not mysterious: they rode, repeatedly, when they otherwise would not have. That is most of what the health benefits of cycling come down to.
If you already train, an e-bike is a logistics tool rather than a training tool. It gets you to work without needing a shower, it lets you ride with your partner or your parent, and it keeps you moving in the fortnight after a crash. Judge it on those terms and it earns its keep. Judge it as a way to get fitter than your training already makes you, and it will disappoint you — which is a reasonable thing for a bike with a motor on it to be honest about.
Sources 9
Where this article summarises a study, the study itself is linked — not a write-up of it.
- 1 Bourne JE, Sauchelli S, Perry R, Page A, Leary S, England C, Cooper AR Health benefits of electrically-assisted cycling: a systematic review · International Journal of Behavioral Nutrition and Physical Activity · 2018
- 2 Höchsmann C, Meister S, Gehrig D, Gordon E, Zhang Y, Nussbaumer M, Rossmeissl A, Schäfer J, Hanssen H, Schmidt-Trucksäss A Effect of E-Bike Versus Bike Commuting on Cardiorespiratory Fitness in Overweight Adults: A 4-Week Randomized Pilot Study · Clinical Journal of Sport Medicine · 2018
- 3 Peterman JE, Morris KL, Kram R, Byrnes WC Pedelecs as a physically active transportation mode · European Journal of Applied Physiology · 2016
- 4 Langford BC, Cherry CR, Bassett DR, Fitzhugh EC, Dhakal N Comparing physical activity of pedal-assist electric bikes with walking and conventional bicycles · Journal of Transport & Health · 2017
- 5 Berntsen S, Malnes L, Langåker A, Bere E Physical activity when riding an electric assisted bicycle · International Journal of Behavioral Nutrition and Physical Activity · 2017
- 6 Sperlich B, Zinner C, Hébert-Losier K, Born DP, Holmberg HC Biomechanical, cardiorespiratory, metabolic and perceived responses to electrically assisted cycling · European Journal of Applied Physiology · 2012
- 7 de Geus B, Kempenaers F, Lataire P, Meeusen R Influence of electrically assisted cycling on physiological parameters in untrained subjects · European Journal of Sport Science · 2013
- 8 Bourne JE, Cooper AR, Kelly P, Kinnear FJ, England C, Leary S, Page A The impact of e-cycling on travel behaviour: A scoping review · Journal of Transport & Health · 2020
- 9 Bull FC, Al-Ansari SS, Biddle S, Borodulin K, Buman MP, Cardon G, Carty C, Chaput JP, Chastin S, Chou R, Dempsey PC, DiPietro L, Ekelund U, Firth J, Friedenreich CM, Garcia L, Gichu M, Jago R, Katzmarzyk PT, Lambert E, Leitzmann M, Milton K, Ortega FB, Ranasinghe C, Stamatakis E, Tiedemann A, Troiano RP, van der Ploeg HP, Wari V, Willumsen JF World Health Organization 2020 guidelines on physical activity and sedentary behaviour · British Journal of Sports Medicine · 2020
See this in your own numbers
Moveee builds your power curve, fitness and fatigue from every ride you upload, and writes a training plan around them. It is free, it takes one Strava connection, and it will tell you more about your riding in a week than any article can.
Start freeKeep reading
A bike maintenance schedule you'll actually keep
Not the aspirational list nobody follows — the realistic one, sorted by how often each job genuinely needs doing. Chain wear measured properly, the cost argument for replacing early, and the three jobs worth paying a shop for.
Bike fit basics: saddle height, setback and reach explained
Three measurements do most of the work, and the research on saddle height is better than you'd expect. What each adjustment changes, how small a change should be, and when to stop guessing and see a fitter.
How to calibrate your power meter (and how often)
A zero offset is not a calibration, and temperature moves your numbers more than most riders realise. What each procedure does, how often to do it, and how to tell a drifting meter from genuinely lost fitness.
10 changes that make you a genuinely faster cyclist
Ten improvements ordered by return on effort — only some of which involve producing more watts. Plus four things you can fix in your garage this afternoon for free.
Disc brake noise: what it means and what to do
Better than rim brakes in nearly every way that matters and worse in exactly one — when they are unhappy they tell you loudly, in front of other people. Diagnosis by sound, the five-minute alignment fix, and the one noise that means stop riding.
Saddle sores: prevention, treatment and when to see a doctor
"Saddle sore" covers several different problems with different treatments, which is why the internet's advice contradicts itself. What is actually going on down there, what genuinely prevents it, and the signs that mean stop riding and get it looked at.
No gaming PC needed: virtual cycling on the laptop you already own
Zwift's system requirements quietly rule out a lot of laptops. A browser-rendered world doesn't. Here's how to get a real virtual ride out of ordinary, years-old hardware.
Your first winter on the bike: kit, roads and staying motivated
Winter is where new riders quietly stop. Warm extremities, lower roads, one hard session a week and a fixed ride day carry more fitness into spring than any jacket.
Commuting by bike: how to make it count as training
Riding to work has some of the strongest health evidence in all of cycling. How to turn thirty minutes each way into genuine training, when to keep it easy instead, and the logistics that make it sustainable.
Should you ride when you're ill? A straight answer
The neck check, the symptoms that mean stop, and how to rebuild afterwards without digging a three-week hole. Plus the reassuring truth about how little fitness a week off costs you.

How to use an indoor cycling app without a smart trainer — a speed sensor is enough
A £30 wheel speed sensor turns any dumb trainer into a real route in a 3D world, with the ride saved to your calendar. What you need, how to set it up, and the one thing about hills you should know before your first session.
Cycling cadence: does it actually matter? What the science says
The internet is full of one true cadence — the research says your legs already found theirs. What the science honestly shows about self-selected cadence, when deliberate high or low RPM work genuinely helps, the typical ranges by situation, and the drills worth folding into an easy hour.