Every e-bike claims a range. Here's how much battery they're assuming per mile.

Published 2026-09-02

Every e-bike claims a range. Here’s how much battery they’re assuming per mile.

Range claims are the least trustworthy number on an e-bike spec sheet. “Up to 60 miles” is doing a lot of work with those two words, and no manufacturer tells you the rider weight, terrain or assist level behind it.

So here’s a different way to look at it. Take the battery capacity in watt-hours, divide by the claimed maximum range, and you get the watt-hours per mile the manufacturer is quietly assuming. It turns a marketing number into an engineering one — and unlike the range claim itself, it can be checked against physics.

We ran it across 86 bikes from 7 brands, every one of them a figure we read off the manufacturer’s own page.

The number

The median e-bike in our database assumes 11.2 watt-hours per claimed mile.

That was also the median when we last ran this over 49 bikes. The sample has since nearly doubled and the figure did not move, which is worth more than the figure itself: it suggests 11.2 is a property of how the industry writes range claims rather than an accident of which bikes we happened to have.

The spread is wide:

To put 11.2 Wh/mi in context: it’s roughly what a 180 lb rider gets on low assist, on flat ground, with no wind. It is not what most people get. If you ride at high assist, into a headwind, up hills, or you’re carrying groceries and a child seat, 15–25 Wh/mi is a more realistic band — which is why the claimed range and the range you actually see differ by so much.

The useful thing about this number is that it’s comparable across bikes in a way range claims are not. A 60-mile claim on a 500 Wh battery and a 60-mile claim on an 800 Wh battery are not the same claim. The first assumes 8.3 Wh/mi; the second assumes 13.3. The first brand is being considerably braver.

The bikes assuming the least battery per mile

These claims lean hardest on ideal conditions:

* capacity computed from the published volts × amp-hours (see methodology).

Two different things are going on in that table, and they deserve different reactions.

The light bikes at the top are doing what light bikes do. The Tempo is 39 lb with a 350 W motor; the CF Racer1 is a 250 W carbon road bike; the C1 is a 32 lb kids’ bike. A light bike genuinely uses less energy per mile, so a low figure here is not dishonest — it’s what that machine should look like. What it does mean is that the range claim assumes you ride the way the bike is happiest: gently, on the flat, at low assist.

The Aventons are a different case. The Level 4 ADV and the Current ADV are 61 lb and 56 lb bikes claiming 110 and 105 miles. Those are three-figure range claims on a bike heavy enough to need real energy to move, and they are what push Aventon to the most optimistic brand median in this dataset. Nothing about them is impossible — 7.3 Wh/mi on a well-geared bike at the lowest assist setting is achievable — but they are claims you should read as a ceiling reached under laboratory conditions, not a planning figure. (The two Level 4 ADV rows are the same bike in a high-step and a step-through frame, which is why they are identical.)

The bike that didn’t make the table

The Velotric T1 would have topped this list at 5.0 Wh/mi. It is excluded, and the reason is more interesting than the number.

5.0 Wh/mi is below what an assisted e-bike can physically achieve, so one of the two figures had to be wrong. Its page publishes 36 V and 9.8 Ah — 352.8 Wh, which matches what we recorded — and claims “up to 70 miles”. But the same page sells a range extender: 36 V, 4.9 Ah, 176.4 Wh. Add it and you get 529 Wh, which across 70 miles is 7.6 Wh/mi — an entirely ordinary figure for a light bike.

So the most likely explanation is that the headline range assumes an accessory that doesn’t come with the bike. We can’t prove that from the page, so we don’t assert it. The row is flagged for a human instead of being published either way.

This is the whole argument for computing Wh/mi in the first place: it is arithmetic on two numbers, so an impossible result means one of them is wrong, without you needing to know which one in advance.

The bikes assuming the most

These are the conservative claims, and one brand owns almost the whole table. The three Revv1 rows are the same platform in three builds — 110 lb moped-style bikes with 1,040 Wh packs claiming 30–60 miles. A bike that heavy really does need that much energy per mile, and quoting 60 rather than 90 is the manufacturer declining to flatter itself.

If you are shopping for a utility, cargo or moto-style bike, this end of the table is the honest end, and a 70 lb cargo bike claiming 8 Wh/mi should make you suspicious rather than pleased.

Note also that we are being generous to these bikes: where a brand quotes a band, we use the top of it. The Revv1’s 17.3 comes from taking its 60-mile figure. Read the bottom of the band instead and it implies 34.7 Wh/mi.

The most useful thing we found

The strongest signal in this data is not in either table above. Lectric sells three of its platforms in both a standard and a long-range version, which means the same bike, twice, with two different batteries and two different range claims. Line them up:

All three pairs imply the same energy consumption for both battery sizes. That is what internal consistency looks like, and it is a real signal about how a company writes its numbers. The alternative — which we do see elsewhere — is a bigger battery quietly buying a disproportionately bigger range claim, where the long-range model implies a lower Wh/mi than the standard one. That can only be true if the bigger pack somehow makes the bike more efficient, which it does not.

You can check this yourself in about two minutes on any brand’s site: find a bike sold with two battery sizes, divide each capacity by its claimed range, and see whether the two numbers agree. If they don’t, the larger claim is the softer one.

By brand

Median Wh per claimed mile, brands with at least three usable models:

Read this carefully, because it is easy to read backwards. A low median is not a better brand — it means that brand’s range claims assume less energy per mile, which means the claims are more optimistic. Ride1Up sits at the conservative end largely because it sells heavy fat-tyre and moto-style bikes; Aventon’s median is pulled down by light commuters and by the three-figure range claims on its ADV models.

The honest summary is that brand-level differences here are mostly a story about what kind of bikes each brand sells, not about who is more truthful. We’d need to compare within a category to say anything stronger, and we will once the database supports it.

What this doesn’t tell you

• It is not a real-range measurement. It’s arithmetic on two manufacturer claims. If the claimed range is fantasy, this figure just tells you how ambitious the fantasy is. Actual range is what the Real Range Database is for, and that needs owner reports.
• “Up to” and a min–max band are different claims. 39 of these 86 bikes state a single “up to” figure; the other 47 give a real number or a band. We use the maximum in both cases, which flatters the “up to” bikes. A bike quoting “20–60 miles” is being more informative than one quoting “up to 60”.
• Battery capacity is not battery quality. Cells, BMS and thermal management differ, and a 720 Wh pack that sags under load will not deliver its rated energy.

Methodology

Every figure is read from the manufacturer’s own published spec page by our scraper, with the source URL and fetch date recorded per model. Nothing here is estimated from a similar model.

What we excluded, and why. Of 113 models in the database:

• 15 are excluded because something on the page contradicts itself and we have not resolved it.
• Of the 98 that remain, 8 publish no battery capacity we can attribute to the specific variant.
• Of the 90 then remaining, 4 publish no range claim at all.

That leaves 86. (The three figures are applied in sequence, so they add up.)

The first of those numbers was 48 when we first ran this analysis. It fell to 15 because our extractor got better at reading particular manufacturers’ pages, not because we lowered the bar — the same checks are applied, and the bikes that still fail them are still excluded. The most common remaining reason is a page that sells one bike in two battery sizes without saying which one the model we’re looking at has; we leave the capacity empty rather than guess.

On variant attribution. Several manufacturers sell one bike in two battery sizes and list both on one page. An earlier version of this analysis paired one variant’s battery with the other variant’s range claim — the Ride1Up Portola 13.4 Ah, for instance, was recorded with the 10.4 Ah model’s 500 Wh and 40-mile figures. Both halves came off the same page, so nothing about the result looked wrong. Every capacity and range in this piece has been re-checked against the specific variant it belongs to, and where the page doesn’t let us tell the variants apart, the model is excluded rather than guessed at.

On plausibility. Any bike whose capacity and range claim imply less than 6 Wh per mile is flagged and excluded, because no assisted e-bike achieves it — the figure means one of the two numbers is wrong. One bike currently trips this (above). The same check catches the mirror-image failure above 40 Wh/mi; nothing in the catalogue is close to that.

On derived capacities. 26 of the 86 bikes publish volts and amp-hours but never state watt-hours. For those we compute Wh = V × Ah, which is the definition of the unit rather than an estimate; they are marked * in the tables above. The medians barely move if you drop them: 11.2 Wh/mi with them, 10.8 Wh/mi without (n = 60).

They are not spread evenly, and that is worth knowing before you weigh any single brand’s number. Ride1Up publishes watt-hours for only 3 of its 13 bikes and Himiway for 3 of 10, while Lectric states them for all 14 and Rad Power for all 7. So the conservative end of our table leans more on arithmetic than the optimistic end does — though since V × Ah is what a watt-hour is, this affects how much we had to compute, not how accurate it is.

On class-switchable bikes. 50 of the 86 ship in one e-bike class and unlock to another. Top speed changes with the mode, and so does real-world consumption — but the manufacturer’s range claim is a single number regardless. See which bikes switch class.

On discontinued bikes. 7 of the 86 are models the manufacturer has since stopped selling — two Aventon Level 2s, two Level 3s, the Ride1Up Rift and 500 Series, and the Heybike Horizon. We keep them in. This piece is about how range claims are written, and a claim does not stop having been made when the bike leaves the catalogue; dropping them would also quietly bias the sample toward whatever each brand is promoting this month. But if you are reading the tables as a shopping list, check the model page — it says whether the bike is still sold.

Data read 2026-08-13 and 2026-08-14, with Aventon re-read on 2026-09-01; figures re-verified against the database on 2026-09-02. Prices and specs change; the model pages carry the date each figure was last verified.

Sources

• Aventon — product pages (17 models in the analysis), read 2026-08-14 to 2026-09-01
• Velotric — product pages (15 models), read 2026-08-14
• Lectric eBikes — product pages (14 models), read 2026-08-14
• Ride1Up — product pages (13 models), read 2026-08-14
• Heybike — product pages (10 models), read 2026-08-14
• Himiway — product pages (10 models), read 2026-08-14
• Rad Power Bikes — product pages (7 models), read 2026-08-14