Why is everyone talking about it this year?

The clearest signal came from search and social rather than sales figures. Trend data presented at the Beauty and Wellness Show in New York in March put LED face mask interest up 34% year on year across TikTok and Google, with Google holding around 70% of that share. Light therapy sat alongside cryotherapy and lymphatic drainage as one of the treatments younger buyers have folded into normal routines.

The market analysts tell the same story more slowly. Grand View Research put the global red light therapy market at around $533.8 million in 2025, rising to roughly $587.5 million in 2026 on a 9.8% growth rate. The interesting detail underneath that number is the split: clinics and spas still account for the biggest share of revenue, but home-use panels and portable devices are now the fastest-growing part of it.

That matches what we see. People try it somewhere else first, then work out that the maths of a panel at home beats paying per session.

What did the regulator actually do in 2026?

The background matters. Late in 2025 the ASA used its AI-driven Active Ad Monitoring system to sweep LED mask advertising and banned ads from several brands for claiming their devices could treat acne and rosacea. The reasoning was procedural rather than scientific: those are medical claims, and a device cannot make them unless it is registered with the MHRA as a medical device. UKCA or CE marking for electrical safety is not the same thing and never has been.

Then in June 2026 the ASA ruled against Beauty Pie over a claim that its LED mask was clinically proven to reduce wrinkles in four weeks. Wrinkle reduction is a cosmetic claim, not a medical one. It was banned because the evidence behind it was judged insufficient, not because the claim was off-limits.

• November 2025

  Four LED mask brands have ads banned for unauthorised medical claims about acne and rosacea.

• Early 2026

  Legal commentary flags the ASA taking a zero-tolerance line on wellness brands straying into MHRA territory.

• June 2026

  Beauty Pie's "clinically proven to reduce wrinkles in four weeks" claim is banned for insufficient evidence.

What did the research find this year?

Volume is not the problem. One review of the literature counted roughly 140 genuine photobiomodulation papers published in July and August alone. The problem has always been that the protocols vary so widely between studies that pooling them is difficult, and 2026 has not fixed that.

What it has done is sharpen one finding that consumers can actually use. Several independent studies this summer, in humans and in cell culture, pointed the same way: there is a therapeutic window, and above it the effect flattens or reverses. In one cell culture line, 9 joules was protective, 30 joules was partially protective, and 60 joules made oxidative stress worse. In human brain imaging work, moderate irradiance raised the measured signal while low irradiance lowered it.

Beyond that, the areas where evidence continued to firm up were the unglamorous ones: musculoskeletal rehabilitation, wound healing and post-operative recovery in clinical settings, using clinical devices under supervision. Those are not home-panel use cases, but they are the reason the underlying mechanism keeps holding up.

Why "more powerful" stopped being a selling point

For most of the last five years the consumer market has competed on a single axis: more LEDs, more watts, bigger numbers. The 2026 literature makes that a poor proxy for results. What matters is the total dose delivered to your skin, which is irradiance multiplied by time, and the useful range for that is narrower than the marketing suggests.

There is a second problem underneath it. Independent verification of irradiance figures across consumer devices is patchy. Published figures for face devices alone span roughly 15 mW/cm² on generic units up to around 60 mW/cm² on premium ones, which is a fourfold difference in how long a sensible session should last. A brand that publishes no figure at all is asking you to guess.

What is still experimental in 2026?

These are the areas producing the headlines this year, and they are the areas where the gap between a research protocol and a consumer device is widest. Studies in these fields use specific devices, specific placements and supervised protocols. A general-purpose panel in your spare room is not the same intervention.

Our position on this is simple. If a seller is using early-stage brain or metabolic research to sell you a panel, they are borrowing credibility from work that does not apply to the product. Buy on skin and recovery, where the evidence actually sits. Treat everything else as a possible bonus that nobody has promised you.

If you are buying in 2026, what has changed?

• Irradiance, with the distance it was measured at. A figure quoted at 5cm will look several times better than the same panel measured at 30cm. Both numbers or neither.
• Both 660nm and 850nm if you want anything beyond skin. Red alone is a skin device.
• A recommended session length. A seller who has thought about dose will tell you how long to use it for. One who has not will tell you it is powerful.
• Claims you can interrogate. After this year, "clinically proven" with nothing behind it is a liability rather than a credential.
• A UK warranty and a UK returns address. Sending a 15kg panel to another continent is not a warranty.

What happens next?

The ASA's monitoring has concentrated on face masks because that is where the paid social advertising volume is. Panels are the obvious next category, and the same rules already apply to them. Sellers who have been sloppy with language have a window to fix it.

The more useful shift is commercial. Once buyers start asking for irradiance figures, the brands that cannot produce them lose their advantage, and the sub-£100 end of the market gets harder to sustain. That would be a good outcome. Most people who conclude red light therapy does not work have simply used a device that never delivered a meaningful dose.

Frequently asked