How we test electric bikes

I am Ravi Kapoor, and I wrench on and ride every bike that gets a verdict on this site. Not a press loaner I rolled around a parking lot for an afternoon. I mean real miles on real roads, real hills, real stop-and-go traffic, in the kind of weather you actually commute in. A spec sheet tells you what a maker wants you to believe. The road tells you the truth, and the road is what I report.

Quickly, so you know what is coming: I ride each bike for weeks, measure how far the battery really goes, take the thing apart enough to check the welds, brakes and cells, then score it on ride quality, build, value and reliability. Prices and specs come straight from each manufacturer and get checked by hand. Nobody pays us for a ranking. The detail behind each of those steps is below.

I ride every bike on real roads, not a spec sheet

Most ebike content online is a spec sheet read out loud. Motor wattage, battery size, top speed, done. That tells you almost nothing about whether you will enjoy the bike on a Tuesday morning with a backpack on. So before I write a word, I put real miles on it.

My test loop has everything a commuter actually hits: flat city blocks with constant stops, a couple of steep climbs that expose a weak motor or a touchy assist, rough chip-seal that rattles loose hardware, and a long flat stretch where I can let the battery drain and watch the numbers. I ride at a real commuter weight with real cargo, not stripped down to game the range.

The single biggest thing I am feeling for is the assist. A torque-sensor bike reads how hard you push and adds power in proportion, while a cadence-sensor bike delivers a preset jolt the moment the pedals move, and that difference does more for how a bike feels than any number on the box (here is the full breakdown). That is why I will tell you the Velotric Discover 2 and the Aventon Level 3, both torque-sensor bikes, feel more natural than the cadence-sensor Ride1Up 700 Series, even though all three run a 750W class motor. The number on the box is the same. The ride is not.

I also lean hard on the brakes on every descent. Hydraulic discs, like the ones on the Lectric XP4 and the 700 Series, should stop a 60-plus-pound bike confidently and stay firm after repeated hard stops. If a lever goes soft or vague, you will read about it.

I measure real range instead of trusting the box

This is where I get the most pushback and where I refuse to budge. The advertised figure is a lab ceiling set up to flatter the bike, so shave a third or more off it for anything resembling a real commute (the full reasoning is here). I do not trust the sticker. I drain each battery on my test loop under normal commuting conditions and log what it actually delivers. A few examples of how that plays out:

• The Lectric XP4 advertises 50 to 85 miles depending on which battery you spec, up to an 840Wh pack. The high number is achievable only on a long-range battery, lowest assist, perfect conditions. Most commuters on the bigger pack will see real-world numbers well inside that window.
• The Velotric Discover 2 claims up to 75 miles from its 706Wh battery, and the Aventon Level 3 claims up to 70 miles from a 733Wh LG-cell pack. Both are honest about being torque-sensor commuters, but a hilly route with the assist turned up will still cut those figures down hard.
• The Ride1Up 700 Series is one of the more upfront here, listing 30 to 50 miles from its 720Wh Samsung battery. That is a realistic spread, and it tends to land where they say.

If you want to understand why your mileage will vary so much, I break it all down in our guide to ebike range. When you shop, ignore the headline number and look at battery watt-hours, then assume real life takes a bite out of it.

I check the welds, brakes, battery and torque by hand

Range and feel are what you notice first. Build quality is what you notice on year three. A bike can ride great out of the box and still be hiding cut corners, so I get hands-on with the parts that matter.

Welds and frame. I look at every weld, especially around the head tube, bottom bracket and rear dropouts where stress concentrates. Clean, consistent beads are a good sign. Lumpy or gappy welds on a 75-plus-pound utility bike like the Rad Power RadRunner Plus are a real concern, because that frame hauls a passenger and cargo.

Brakes. Beyond riding them, I pull the wheels, check the rotors for true and the pads for even contact, and bleed-test the hydraulic lines for sponginess. Soft brakes are the most common safety shortcut on cheap ebikes, and I call them out every time.

Battery and cells. The battery is the most expensive and most safety-critical part of the bike. I check that the pack seats and locks solidly, that the charge port is clean, and crucially whether the cells and pack carry real safety certification. The Velotric Discover 2 uses a UL-certified battery, and the Level 3 runs known-quantity LG cells. That matters more than a few extra watt-hours.

Torque and fasteners. I put a torque wrench on the stem, headset, brake mounts and motor bolts, because shops sometimes ship bikes loose and a rattling fastener becomes a failed part. I note anything that arrived under-tightened so you know to check it before your first ride.

How I score: ride, build, value and reliability

Every bike gets judged on four things, and I weigh them in roughly this order because that is how they actually affect your daily life with the bike.

No single spec wins on its own. A bike with a torque sensor, certified battery and clean welds will beat a cheaper one on paper even if the cheaper one claims more range, because day to day the better-built bike is the one you keep riding. If you want to see how I apply this to actual buying decisions, start with our guide to buying an electric bike and the roundup of the best commuter electric bikes.

Where specs and prices come from

I do not copy numbers from other review sites, because errors get repeated and multiplied that way. Every motor wattage, battery capacity, weight, range claim and price you see here is pulled from the manufacturer directly, then checked by hand against the current product page before it goes live.

I am careful to keep two different numbers separate. There is the advertised figure, which is the maker's claim, and there is my real-world result from riding the bike. When you read that the RadRunner Plus is rated for 55-plus miles on its 624Wh battery, or that the Aventure 3 makes 750W with a 1,440W peak, those are the manufacturer's specs as listed. My measured numbers and my opinion are clearly mine. Prices move, so if you are checking current pricing on a bike like the Lectric XP4 or the Velotric Discover 2, the maker's own page is the source of truth, and I keep ours synced to it.

How we stay independent

Watt & Wheel makes money through affiliate links. When you buy a bike after clicking through to a brand like Aventon or Ride1Up, we may earn a commission at no extra cost to you. That is the whole business, and I want to be plain about it.

What that commission does not buy is a soft review. I have recommended cheaper bikes over pricier ones plenty of times because the cheaper one was simply better for the rider, and the payout had nothing to do with which way I leaned. If a bike has a soft brake or an optimistic range claim, I say so whether or not there is a commission attached to it.

You can read the full affiliate disclosure, learn more about who we are, or check my own background on the author page. If you ever think a review reads like an ad, call me on it, because that is exactly the kind of content I built this site to replace.