Most red light panel specs are marketing. We measure them instead.

Manufacturers quote irradiance at distances they do not name, using meters that read near-infrared badly. We test at fixed distances, publish the method, and show our working.

Last checked

There are two numbers that decide whether a red light panel is worth buying, and almost nobody publishes either one honestly.

The first is irradiance at a stated distance — how much light actually reaches your skin, in milliwatts per square centimetre, measured somewhere you would realistically stand. A panel quoting 200 mW/cm² without naming a distance is quoting a number taken with the sensor pressed against the lens. That figure is true and useless.

The second is what it costs to get it into the UK. Most of the panels worth owning ship from North America, and the sticker price is the smaller half of the bill. Import VAT, duty and a courier handling fee routinely add a third.

What we publish

  • Measured irradiance at 15cm, 30cm and 45cm, sensor centred and off-axis.
  • The spectrum, so you can see what is actually in there rather than what the box claims.
  • Landed UK cost, worked through line by line, in pounds.
  • The method, in enough detail that you could repeat it and catch us out.
380nm — 950nm. Dashed markers are near-infrared: they do real work and emit no visible light, which is why a panel’s brightness tells you almost nothing about its output.

Where we are up to

This site is new and the bench work is in progress. Rather than pad the rankings with numbers we have not taken, pages state plainly what has been measured and what has not. A panel appears in the table when it has been through the test, not before.

The honest caveat

Red light therapy is an area where the marketing has comfortably outrun the evidence. The physics of dose is well defined and we treat it as such. The clinical claims are a different matter: much of the literature is small, short and industry-adjacent. We measure devices. We do not tell you what a device will do for you, because on most endpoints nobody yet reliably knows.