The Best Helmets for E-Bike and PEV Riders

Most "top rated" helmet lists are quoting scores that no longer exist

Virginia Tech runs the only large independent helmet testing programme in the US. Twenty-four impact tests per helmet, a public score, a star rating. It is the best evidence a buyer has.

In July 2025 they reset the thresholds, and 139 helmets lost their 5-star rating. Helmets had genuinely improved, so the old scale had stopped separating anything: nearly everything mainstream was 5 stars. The new scale spreads the tested helmets across the full range: of the 311 on the list today, only 54 are still 5 stars.

Almost nobody updated. Brand pages, buying guides and the AI-written listicles that fill this search result are still repeating pre-2025 numbers, which is why you will see helmets described as "5-star rated" that now score 2. Every rating on this page was pulled directly from Virginia Tech in September 2026 and reflects the current scale. Every pick holds 4 or 5 stars, the level Virginia Tech itself recommends, and the 5-star helmets come first.

Two honest caveats. A low score does not mean a helmet is unsafe: every helmet here passes CPSC, which is legally required. Virginia Tech measures how far past that floor a helmet goes, so the real distinction is "meets the standard" against "measurably better than the standard". And it is one lab with one protocol, which is the best public evidence available rather than the last word.

What is MIPS?

This is the question new riders ask most, so here is the short version.

When you fall off, you rarely hit the ground straight on. You hit it at an angle. Your head stops, but it also twists, and that twisting is what causes concussion. A plain helmet handles the stopping and does very little about the twisting.

MIPS stands for Multi-directional Impact Protection System. It is a thin, low-friction liner inside the helmet that lets the shell rotate about 10 to 15 millimetres against your head during an impact, shedding some of the twist before it reaches your brain.

That is the whole idea. It adds a few grams and, these days, very little money.

Do you need to hunt for MIPS specifically? No. Several makers built their own versions: WaveCel (Trek and Bontrager), KinetiCore (Lazer), SPIN (POC), 360 Turbine (Leatt). The current expert position is that these all work and that no ranking between them is supportable. Virginia Tech's 5-star list contains MIPS, WaveCel and KinetiCore helmets side by side. So: have one of them, and stop worrying about which.

If the same helmet is sold with and without a rotational system, take the version that has one.

The certifications, and which ones actually mean something

CPSC 1203 is the floor, not a ranking. Every helmet legally sold as a bicycle helmet in the US meets it. It is pass or fail, tests a single impact, and tells you the helmet cleared the bar and nothing about how far it cleared it. Two helmets with identical stickers can perform very differently.

NTA 8776 is the one that matters most for PEVs. It is a Dutch standard written for speed pedelecs: tested to 45 km/h (28 mph), with more tested shell area at the temples and the back of the head than a bicycle standard demands. If you ride a Class 3, this is the single most relevant thing on the label. US distribution is thin, so its absence is not a red flag, but do not assume CPSC covers 28 mph riding. It was not written for it.

ASTM F1492 is the skate standard, and it is not the answer it looks like. CPSC tests one hard impact, on the assumption that you crash once and replace the helmet. F1492 tests multiple impacts, because skaters fall repeatedly in a session. That sounds like a match for scooter, one-wheel and e-skate riders, who tend to go down at lower speed but more often. The measured numbers say otherwise: the three dual-certified skate helmets Virginia Tech has tested score 1, 1 and 0 stars, and the two skate-shaped helmets that add NTA 8776 manage 3. Virginia Tech's test is built around cycling impacts and skate shells are not designed for it, which explains the gap without closing it. So we do not recommend a skate helmet here. If you ride a scooter, one-wheel or electric unicycle, look at a full-face instead: our full-face helmet guide compares the ones that score well.

ASTM F1952 is the downhill standard, and it is the only one that tests a chin bar. If you are looking at a full-face helmet, this is the certification to look for. More on that below.

EN 1078 is the European equivalent of CPSC. Lower drop heights, but a stricter g-force limit. Seeing it alongside CPSC is a mild positive, not a decisive one.

Snell B-95 subjects helmets to harder hits than CPSC over a larger area, but it is voluntary and very few helmets carry it.

Getting the fit right, which decides whether any of this works

Experienced riders say this before they say anything else, and they are right. A helmet that moves on your head redistributes an impact somewhere useless. One that pinches gets left at home. Neither protects you.

Measure first. Wrap a soft tape around your head about a centimetre above your eyebrows, at the widest point. That number in centimetres is what the size chart wants. Do not guess from hat size.

Then check four things on your head:

1. It sits level, with the front edge about two finger-widths above your eyebrows. Tipped back is the most common mistake and it leaves your forehead bare.
2. The side straps form a Y just under each ear, with the join sitting right below the earlobe.
3. You can slide two fingers under the chin strap and no more.
4. Shake your head hard. Nod, then side to side. The helmet should move your scalp with it. If the helmet slides independently, the size or the shell shape is wrong.

That last one is the real test. Every brand's shell shape is different, and pad swapping cannot fix a shell that is the wrong shape for your head. If you can try helmets on in person before buying, do that, even if you order online afterwards.

Match the coverage to how your vehicle throws you

A vented road helmet stops well above the ears and leaves the back of your skull comparatively exposed. Urban and skate shells drop lower at the rear and sides. At 15 mph that difference is academic. At 28 mph, or on a one-wheel where a rearward fall is the classic crash, it stops being academic.

• Class 1 and 2 e-bikes, mostly pedalling: a good vented helmet is fine. Prioritise comfort so you keep wearing it.
• Class 3 e-bikes, 28 mph: look for NTA 8776 first, coverage second.
• Scooters under about 20 mph: any helmet on this page. Pick an urban shape for its lower coverage at the back, and skip skate helmets, which score poorly when measured (see above).
• One-wheels and electric unicycles: a full-face helmet. Our full-face helmet guide covers the ones that score well.
• Anything above about 25 mph: a chin bar, whatever you ride. See below.

Match the shape to how your vehicle actually throws you, not to what looks fastest.

Ventilation is a real trade, not a luxury

More vents means a cooler head and a lighter helmet, and less shell material. On an e-bike you are pedalling less and moving faster, so airflow is higher and your own heat output is lower than on an acoustic bike. That tilts the balance toward more coverage than a road cyclist would choose. If your commute is short, prioritise coverage. If you are climbing in summer heat, you will want the vents.

Lights, turn signals and smart helmets

A light on your helmet sits at the highest point on you, moves where you look, and is never the thing you left on the kitchen table. That is a genuine advantage over a bar-mounted light, not a gimmick.

Turn signals used to mean giving up measured performance, because the helmets that had them scored poorly. That changed: the UNIT 1 Aura holds a current 5-star rating with signals, a brake light and crash detection built in. If you want the electronics, you no longer have to trade safety for them.

Two warnings.

Bluetooth intercom helmets are a separate category. Sena's cycling helmets are the ones riders mention most, and they are genuinely good at what they do. But they are certified to CPSC 1203 and EN 1078 only, are not NTA 8776, and have not been tested by Virginia Tech. If group comms is the thing you want most, that is a legitimate choice, but know that you are buying an unmeasured shell.

Check which model you are actually looking at. Lumos, for example, sells several helmets under similar names: the plain Lumos Ultra scores 2 stars and is CPSC only, while separate Ultra E-Bike and Sonorus models carry NTA 8776. The names are one word apart and the specifications are not the same.

The cheap smart helmets you will find on Amazon

If you search for an e-bike helmet you will land on a wall of $40 to $90 helmets with LEDs, visors and rechargeable tail lights, from brands you have never heard of. Some are legitimately CPSC certified. The question is what the certification is worth.

Virginia Tech has tested a few of them, and the answer is: not much. Victgoal VG 112 scores 2 stars. Base Camp BC-019 and the Base Camp Urban Commuter both score zero. ILM's E3-12LS manages 2 stars even though it carries NTA 8776 certification.

That is not an argument for riding bare-headed. Any CPSC helmet beats none. It is an argument that the LED-and-visor feature list on a cheap helmet tells you nothing about how it performs, and that price is not the problem. The P2R A20 MIPS is also an unfamiliar brand selling a helmet with a rear light on Amazon, for about $60, and Virginia Tech ranks it 10th of all 311 helmets tested. Look the exact model up before you buy.

Lock ports and crash replacement

These sound like small conveniences and turn out to change behaviour. A helmet you can lock to the vehicle is a helmet that is with you when you need it, rather than left at home because carrying it around a shop is annoying. Thousand's PopLock is the best known; several makers now do a version.

Separately, a number of brands run crash-replacement programmes that discount a new helmet after a documented crash. Worth checking before you buy rather than after.

When to actually replace a helmet

You will read everywhere that a helmet expires after five years. That is a manufacturer line, and the evidence does not support it.

The Bicycle Helmet Safety Institute tested 675 used helmets, some of them 26 years old, and found that the EPS foam liner does not deteriorate with age. Sweat does not degrade it either: the standards do not permit a helmet that would degrade from sweat, and the foam is largely unaffected by salt water.

Here is what actually means replace it:

• Any real impact. Foam crushes once. A helmet that has taken a solid hit has spent its protection even when the shell looks fine. Multi-impact ASTM helmets are the only partial exception.
• Visible damage. Cracked or crushed foam anywhere. Crushed EPS springs back a little, so a dent that looks minor may not be.
• UV damage to the shell. Fading, or fine cracks around the vents, means the UV inhibitors are failing. This is the one that genuinely ages a helmet, and it is driven by sun exposure, not by the calendar.

Five years is a reasonable interval to think about upgrading, because helmets keep improving and a five-year-old helmet is probably beaten by something current. That is a different claim from "your helmet has expired", and the difference matters if the five-year line is what is stopping you from buying a good helmet today.

Do you need a full-face?

For some PEV riders, yes, and it is under-discussed. If you ride a one-wheel or electric unicycle, or a scooter you regularly take above 25 mph, this is where we would start, rather than with any helmet on this page. If you ride a scooter at commuting speed, an open-face helmet from this page is a sound choice, and the full-face guide is there if you want the extra coverage. Roughly half of electric-unicycle injuries involve the head or face, and no amount of star rating on an open-face helmet protects a jaw. The rule of thumb the EUC community uses travels well to fast scooters: under 20 mph a good bicycle helmet, 20 to 28 mph a downhill or NTA 8776 helmet, above 28 mph a motorcycle-certified helmet.

That is a whole separate decision with its own certifications, and it turns on one fact most guides miss: neither CPSC nor EN 1078 tests a chin bar at all. ASTM F1952 is the standard that does. If you are shopping for a full-face, that is the sticker to look for. Our full-face helmet guide compares the ones worth buying, with their Virginia Tech ratings.