Bike light lumens explained

Heads up: we earn a commission if you buy through the links on this page, at no extra cost to you. It never decides what makes the list — here’s how that works.

Here is the fact this page is built around, and you can check both halves of it yourself.

Every bike light sold advertises lumens. Not one lighting regulation we could find uses the unit.

Germany’s road traffic regulation governs bicycle lamps by beam geometry and dazzle. The US federal bicycle standard governs reflectors and specifies no lamp at all. Between them they cover a lot of the world’s cycling, and lumens appears in neither. Meanwhile the number is on the front of every box, in every product name, and at the top of every roundup — including ours, because it’s what the sellers publish and we’re not going to invent an alternative.

So what’s going on? Nothing conspiratorial. Lumens is simply the easiest thing to measure, the easiest thing to make bigger, and the easiest thing to print. It is also, for the question you actually care about, close to useless on its own. Here’s why.

Lumens measures the lamp. Lux measures the road.

A lumen is a unit of luminous flux: the total quantity of visible light a source emits, added up over every direction it goes. It is a property of the lamp, in isolation, measured as if you had enclosed it in a sphere and counted everything that came out.

A lux is a unit of illuminance: how much light actually lands on a surface, per unit of area. One lux is one lumen spread over one square metre. It is a property of the lamp and the optics and the distance and the angle — which is to say, a property of the situation you are actually in.

That definitional difference has a consequence that decides everything. Because lumens counts light in every direction, a lamp scores identically whether it puts its output on the road in front of you or throws it into the sky, the hedge and an oncoming rider’s eyes. Take a 1000-lumen lamp and spread it evenly over a huge area and you get a dim, useless wash that still says 1000 lumens on the box. Take the same 1000 lumens and focus it into a tight, well-shaped pool on the tarmac where you need it, and you get a genuinely good light that also says 1000 lumens on the box.

The number cannot distinguish those two lights. It is not that lumens lies. It is that it is answering a different question from the one you asked.

The regulator that actually cares: StVZO

Germany regulates bicycle lighting seriously, and its rules are public. Under StVZO § 67, a bicycle must carry “one or two forward-facing headlights for white dippedbeam” and — the operative clause — the headlight “must be adjusted so that it does not dazzle other road users”. Flashing headlights are prohibited. The regulation also sets mounting heights: 400–1,200 mm for the headlight, 250–1,200 mm for the rear light.

Read what that regulation is about. It is about where the light goes. Dipped beam, no dazzle, mounted in a specified band of heights. It is a rule about geometry, written by people whose actual problem is that a bright light aimed at a windscreen makes a road less safe, not more.

Lupine, a manufacturer that builds StVZO-approved lamps and therefore has to satisfy this rule, spells out the consequence in its own explainer: “A common misconception lies in the evaluation of brightness. While many lamps prioritize the lumen count, the lux count is crucial for StVZO-compliant lighting. Lux describes how much light actually reaches the road.” It also describes the compliance mark: the “K-number”, which “indicates that the lighting system has been tested and complies with legal requirements”, and notes that “a precise light-dark boundary is mandatory to avoid dazzling other road users”.

Weigh who is speaking.Lupine sells lights, and it sells lights engineered for a standard it is here telling you is the right standard. That is a commercial interest and you should hold it in mind. But the regulation it’s describing is a public text we linked above and read ourselves, and the physics under the claim is checkable arithmetic rather than opinion. A seller with an interest can still be right; it just doesn’t get to be the only witness.

The cut-off line, and why it’s the real spec

The thing StVZO effectively forces manufacturers to build is a cut-off— a shaped lens that produces a defined light-dark boundary, so the beam illuminates the road ahead and stops before it reaches the eye-line of anyone coming the other way. If you’ve seen a car’s dipped headlights against a wall, you’ve seen one: a bright field with a hard horizontal edge across the top.

This is the single largest quality difference between bicycle lights, and it is almost never published. A light with a cut-off can run more usable light on the road than a light without one, at a lower lumen figure, while blinding nobody — because none of its output is being wasted upwards into the faces of other road users. A light without a cut-off is a torch with a bracket. It solves your problem by creating one for everyone else.

For a sense of what the useful range even looks like in the unit that matters, bike-components’ lighting guide publishes lux recommendations by conditions — roughly 30–60 lux for city and well-lit paths, above 70 lux for unlit roads, above 100 lux for unpaved trails. That is a retailer publishing guidance, not a standards body, and we flag it as such; we quote it because it is one of very few places putting numbers on the right unit at all.

Now the American half, which is stranger

The US federal regulation that governs bicycles as consumer products, 16 CFR § 1512.16, requires an “essentially colorless front-facing reflector”, a “red rear-facing reflector”, reflectors on the front and rear surfaces of the pedals, and side reflectors — all so a bicycle can be recognised “under illumination from motor vehicle headlamps”.

Sit with the implication. The federal standard for bicycles in America has no lamp requirement at all.It is a standard about being lit up by someone else’s headlights. It regulates reflectivity, not light. So the US has no federal opinion on lumens because the US has no federal opinion on bicycle lamps.

Whether you must run an actual light at night where you live is a matter of state traffic law, which we have not audited across fifty states and will not summarise from memory. Check your state. What we can tell you is that the rulebook governing the bike itself talks about reflectors, and that neither of the two regimes on this page has ever asked a manufacturer for a lumen figure.

So what should you actually look at?

The frustrating, honest answer is that the specs you want are mostly not published — which we demonstrate on the lights roundup, where six listings between them publish zero lux figures, zero cut-off claims and zero runtimes. Given that, the useful moves are:

Decide if you’re seeing or being seen, because it changes which spec even matters. Being seen is a question about how visible your light is off-axis — from the side, from a junction, in daylight. Seeing is a question about beam shape and lux on the road. A light optimised for one is usually mediocre at the other, and the lumen figure speaks to neither.

Look for a K-number if you want a light with a proper beam, even outside Germany. It is the only widely available mark anywhere that certifies a bicycle headlight has been tested for its beam pattern and dazzle rather than its raw output. You are borrowing another country’s regulator to do quality control the market otherwise won’t.

Aim the thing. This is free, takes thirty seconds, and matters more than the difference between any two lights on the market. A well-aimed 400-lumen lamp lights the road. A badly-aimed 1200-lumen lamp lights the trees and the eyes of everyone coming towards you, while lighting less of the road, because most of its output is going somewhere useless. Point it down until the bright part lands on the tarmac maybe ten to twenty metres ahead rather than at the horizon.

Treat the lumen figure as an upper bound on a mode you may never use. The number on the box is the maximum output, in boost, for an unpublished period. It is not a description of how the light behaves on your commute.

What we haven’t done

We have not measured a lux figure, photographed a beam, or run a light until it died. There are no beam shots here because we have no lights and no light meter, and a beam shot is exactly the kind of thing that’s easy to fake and hard to verify. Everything above is either a public regulation we fetched and quoted, a manufacturer’s own words with its commercial interest flagged, a retailer’s published guidance flagged as such, or a definition you can check in any physics text. Where we don’t know, this page says so — and on the specs that actually matter for buying a light, we mostly don’t, because nobody publishes them.