Lumens vs Watts: One Number Is Not Brightness
Home
Yiwu, Zhejiang, China
Mon – Fri 9:00 AM – 6:00 PM · Sat 9:00 AM – 4:00 PM
[email protected]

Lumens vs Watts: Why One Number No Longer Means Brightness

Alan Xiao, Founder of Sourcing Pioneer
Alan Xiao
· Updated November 6, 2026 · 9 min read

A watt measures power going in. A lumen measures light coming out. They are not two scales for the same thing, so there is no fixed conversion between them. The number that lets you compare two lamps is lumens per watt. That is why an 8 W LED can replace a 60 W bulb, and why an 8 W lamp from a weaker factory cannot.

Every conversion table you have seen assumes a ratio. Those tables were built for one technology, and that technology is no longer the one being sold.

I get asked for “the lumens of a 50 W” several times a month. The honest answer is that the watt figure tells you the electricity bill and nothing about the light.

This page gives the efficiency of each lamp type and the ceiling that catches impossible claims. Then the difference between lumens and lux, and the numbers to put on an order.

Warm beige studio still life of an LED lamp and an incandescent bulb side by side
Same light on the wall. One of them draws about a seventh of the power.

The Short Answer

If you are choosing a lamp for a product, compare lumens per watt first. That single figure tells you how much light the buyer gets for the electricity.

If you are matching an old incandescent, work backwards. Take the lumen output the old lamp produced, then pick a new lamp at that output. Do not start from the wattage.

In numbers: a 100 W tungsten lamp at 230 V produces 13.8 lumens per watt, so roughly 1,380 lumens. A modern LED screw-base lamp reaches about 102 lumens per watt. To match 1,380 lumens it needs about 13.5 W.

That is the whole subject in one line. The same light, at roughly one seventh of the power, and the watt figure no longer describes the light at all.

Why the Watt Number Stopped Working

For most of the last century, a household had one lamp technology and one efficiency range. Wattage was a fair stand-in for brightness because everyone was buying the same physics.

Tungsten was never efficient, and it got worse as it got smaller. A 15 W tungsten lamp at 230 V manages 8.0 lumens per watt. The 40 W version reaches 10.4, and the 100 W version reaches 13.8.

That last line matters. The efficiency of an incandescent lamp changes with its wattage, because a hotter filament wastes less heat. So even inside one technology, there was never a single multiplier.

Halogen improved the ratio a little. A 100 W halogen at 230 V reaches 16.7 lumens per watt, a 500 W version reaches 19.8, and a low-voltage quartz halogen reaches 24.

Then LED arrived and broke the pattern completely. A 5 to 16 W LED lamp at 230 V lands anywhere from 75 to 217 lumens per watt. What decides where is the chip and the driver inside it.

A 15 W LED at the low end and a 15 W LED at the high end differ by nearly three times in light output. The watt figure is identical. So a table that says “15 W equals 1,500 lumens” is wrong for most products on the shelf.

Lamp typeEfficacy, lumens per wattWhat a buyer notices
Candle0.1 to 0.2Almost nothing, useful as a benchmark
Tungsten, 15 W at 230 V8.0Warm, dim, cheap
Tungsten, 100 W at 230 V13.8The old household standard
Tungsten, 100 W at 120 V17.5Higher, because of the lower mains voltage
Halogen, 100 W at 230 V16.7Slightly better than tungsten
Quartz halogen, 12 to 24 V24Common in small spots
Compact fluorescent46 to 75Good output, slow to warm up
T8 tube, electronic ballast80 to 100The office standard for decades
LED screw-base lamp75 to 217The range is the problem
Bar chart of luminous efficacy in lumens per watt for candle, tungsten, halogen, fluorescent and LED lamps
Candle to LED is more than a thousandfold. No single multiplier spans that.

The table is the reason every conversion chart fails. It spans more than two thousand times from candle to LED, and the upper range overlaps the middle. Every figure comes from the published efficacy table, not from a chart built by a retailer.

The Ceiling That Catches an Impossible Claim

There is a hard physical limit, and it is useful when a supplier sends a number that looks too good.

The eye is most sensitive to light at 555 nanometres, which is green. At that exact wavelength, one watt of light produces 683 lumens. That is the maximum possible, and no lamp can beat it.

White light cannot use that wavelength alone. To look white, a source has to mix red and blue, which the eye sees far less efficiently. So every white lamp sits well below 683 lumens per watt.

The gap is large. A typical incandescent lamp converts only about 2 percent of its input into visible light. The rest leaves as heat.

For white LEDs, the theoretical limit from phosphor colour mixing sits around 260 to 300 lumens per watt. That is the number to remember.

So when a quote claims 400 lumens per watt for a white lamp, it is not describing a better factory. It is describing a measurement taken somewhere other than the lamp’s own terminals.

Two honest explanations exist for such a figure. The supplier may have quoted the LED chip alone, measured before the driver and the optics take their share. Or the supplier may have measured a single laboratory sample rather than a production unit.

Ask which of the two it is. The answer tells you whether the number can survive a shipment.

Lumens, Lux and Candela

Three words get used as if they were one. They measure different things, and mixing them is how a bright lamp ends up lighting nothing.

UnitWhat it measuresDefinition
LumenTotal light leaving the lampLuminous flux, candela times steradian
CandelaLight in one directionLumens per steradian
LuxLight landing on a surfaceOne lumen per square metre

The lumen figure on a box is the total, gathered from every direction in a measuring sphere. It says nothing about where that light goes.

Lux is what a person actually experiences. One thousand lumens spread evenly over one square metre gives 1,000 lux. The same thousand lumens spread over ten square metres gives 100 lux.

Same lamp, same lumens, ten times less light on any given patch of floor. That is why a lumen figure alone cannot tell you whether a work light will be usable.

For a guide, the useful reference points are rough but well known. An office runs at 320 to 500 lux. A hallway sits near 80, a railway platform near 150, and an overcast day reaches about 1,000.

A factory floor is specified in lux at the working plane, not in lumens at the fixture. That is the language the customer’s safety officer will use.

Diagram of one thousand lumens falling on one square metre and on ten square metres with the resulting lux values
The lamp did not change. The area did. Lux is the number that notices.

What Lands on the Floor

Beam angle decides how the same lumens are shared out, and it is the line most order sheets leave blank.

A narrow beam pushes the light into a small solid angle, so the intensity in that direction is high. A wide beam spreads the same total over a much larger angle.

Here is the arithmetic, done for an ideal case. Take 1,000 lumens and put it in a 15 degree cone, spread evenly. The solid angle is about 0.054 steradians, which gives roughly 18,600 candela.

At three metres that is about 2,070 lux. Put the same 1,000 lumens in a 120 degree cone and the intensity drops to about 318 candela. At three metres it is about 35 lux.

Same lamp, same watts, same lumens, and a difference of nearly sixty times in the light on the floor. That is the gap between a spotlight and a floodlight, and it is invisible on most packaging.

Two more effects follow from geometry. Illuminance falls with the square of the distance, so doubling the height cuts it to a quarter. It also falls as the surface tilts away from the beam, by the cosine of the angle.

This is why a street light or a flood light is quoted in lux at a stated distance and height. The solar street light notes work through that specification, because a pole height chosen badly wastes half the output.

Lumens at Hour Zero, and at Hour 25,000

A lamp is measured when it is new. It is used when it is old, and LED output falls with age.

The industry describes that fall with an L number. L70 means the point where output has fallen to 70 percent of the starting figure. L80 means 80 percent.

Two standards sit behind the claim. LM-80 is the test that measures how a family of LED packages holds its output over thousands of hours. TM-21 is the method that projects forward from that data.

So a claim of “L70 at 25,000 hours” is a projection from a shorter test, not a measurement taken at 25,000 hours. That is normal and legitimate. It becomes a problem when the projection is based on a different chip from the one in the box.

For a buyer, the question is narrow. Ask for the LM-80 report for the exact part number, and ask which hour the projection reaches L70. If the report names another part, the number does not belong to your product.

Two other derating effects push the real figure down. Heat inside a small housing shortens life, and a driver running near its limit does the same. A 100 lumen per watt lamp in a sealed fitting is not a 100 lumen per watt product.

Six Lines for the Order

A watt figure is not a specification. These six lines turn a lamp into something a factory can be held to.

  1. Flux in lumens, with the test condition. State the lumen figure and the sphere or goniometer it was measured in, at a stated ambient temperature.
  2. Efficacy, and where it was measured. Write lumens per watt and say whether that is the lamp alone or the complete product with its driver and optics.
  3. Beam angle, or lux at a distance. For a directional product, give the angle. For a flood or street light, give lux at a stated height and distance.
  4. Colour point and colour rendering. Name the colour temperature and the rendering index, and say which bin is acceptable from shipment to shipment.
  5. Lumen maintenance. Name the L value, the projected hour, and the part number the LM-80 report covers.
  6. Power side. Give rated wattage, input voltage range and power factor. A driver that meets its wattage on paper may not meet it at low mains voltage.

Line three is the one that decides whether a customer is satisfied. A product that hits its lumen target in a narrow beam is not the same product as one that hits it in a wide beam. Both carry the same number on the box.

Work lights make this concrete, because they are judged on the pool of light they throw rather than on the fixture. The work light sourcing notes cover how that pool is specified.

Light bars add a second variable, since the output is spread along a length rather than from a point. The light bar ordering notes cover the length and spacing side of it.

How to Compare Two Quotes

Three steps settle most of these comparisons, and none of them needs a laboratory.

First, put both quotes in lumens per watt at the same measurement point. If one figure covers the lamp and the other covers the complete product, the comparison is meaningless until they are aligned.

Second, compare the beam angle. A quote with a tighter beam and the same lumens delivers more light where it is aimed. If the application needs a wide wash, a tight beam is the wrong product regardless of its efficiency.

Third, compare the L70 projection and the part number behind it. Two products with the same starting output and different maintenance figures are not the same product after a year of nightly use.

Only after those three do price and warranty belong in the conversation. A cheaper lamp at a lower maintained output costs more per delivered lumen-year, and that is the figure the buyer’s finance team will eventually compute.

One last check belongs on the test bench rather than the spreadsheet. Ask for the third-party report that covers the finished product, not only the chip. The test file for a finished device is where a claim that survived the quote usually fails.

THE LUMEN LINE

Send the Number, Not the Wattage

Send me the lumen target, the distance it must reach and the area it must cover. I will tell you what efficacy that implies and whether the beam angle on the drawing can deliver it. I will also say which figure on the quote you can hold the factory to. That answer costs nothing before tooling.

Send the Target
How many lumens is a 60 W bulb?

About 800 for a tungsten lamp at 230 V, since 100 W reaches 13.8 lumens per watt and smaller lamps do worse. There is no fixed figure, because efficiency changes with wattage.

Is a higher wattage always brighter?

No. Wattage is power going in. Two 15 W lamps can differ by nearly three times in output, because efficiency depends on the chip, the driver and the optics.

Can lumens be converted to watts?

Only by assuming an efficiency, and the assumption belongs to one lamp type. A table built for tungsten is wrong for LED by a factor of seven or more.

What counts as a good lumens per watt figure?

It depends on the technology. Tungsten sits near 14, halogen near 20, fluorescent tubes near 90, and LED lamps range from 75 to 217. For white light, anything above about 300 is not credible.

Why does a 1,000 lumen lamp look dim?

Usually the beam and the distance. One thousand lumens in a narrow cone is several thousand lux at three metres. The same lumens spread over a wide cone is a few dozen lux.

Does colour temperature change the brightness?

It changes how bright a room looks, not how much light is emitted. Lumens stay the same while the sensation shifts. Perceived brightness is a separate subject from measured output.

Questions & Comments

We read every one and reply within 24 hours

Lumens tell you what leaves the lamp. Lux tells you what lands on the floor. Send me the distance and the area, and I will tell you which figure to hold the factory to.

Subscribe
Notify of
guest
0 Comments
Most Voted
Newest Oldest
0
Would love your thoughts, please comment.x
()
x