Does Red Light Therapy Work? What the research Says (2026)

What is red light therapy?

Red light therapy (also called photobiomodulation or low-level light therapy) shines low-power red or near-infrared light onto the skin. It does not heat or burn the skin like a laser used for resurfacing, and it is not the same as UV tanning. The leading theory is that the light is absorbed by an enzyme in mitochondria (the energy-producing parts of cells) called cytochrome c oxidase, nudging cells to make more energy and calming inflammation (Avci et al., 2013).

Which wavelengths matter?

• Red light, about 630 to 670 nm (nanometres): visible red. Absorbed mostly in the outer few millimetres of skin, so it is the main wavelength for skin and scalp devices.
• Near-infrared, about 810 to 850 nm: invisible to the eye and reaches deeper tissue, which is why it is used for joints and muscles.
• Some masks add 1,064 to 1,072 nm "deep near-infrared". Evidence for this band is thinner.

Blue light (around 415 nm) is a different tool: it targets acne bacteria rather than tissue repair.

Why dose matters more than the label

Two numbers decide whether a device can match the studies:

• Irradiance (mW/cm2): how much light power lands on each square centimetre of skin. It falls sharply as you move the device away.
• Fluence or dose (J/cm2): irradiance multiplied by time. A mask giving 30 mW/cm2 for 10 minutes delivers about 18 J/cm2 (30 x 600 seconds / 1,000).

More is not better. Light therapy has a "biphasic" response: too little does nothing and too much can cancel the benefit (Avci et al., 2013). A device that hides its wavelength or irradiance gives you no way to judge its dose.

What it claims

Social media and brand pages credit red light with smoother, younger-looking skin, clearer acne, hair regrowth, faster healing, less joint pain, quicker muscle recovery, better sleep, fat loss, a better-working thyroid and higher testosterone. Some of these claims are grounded in decent trials; others borrow credibility from clinic lasers or from cell and animal studies.

What the evidence shows

Skin ageing

This is the most consumer-relevant claim with real support. In a controlled trial, people treated with red light had smoother skin and higher collagen density on ultrasound than untreated controls (Wunsch and Matuschka, 2014). A meta-analysis of 31 LED studies found consistent results for skin rejuvenation (Ngoc et al., 2023). The American Academy of Dermatology describes results as "subtle to noticeable" and notes that devices, doses and study designs vary too much to compare directly (AAD). Expect modest change after 8 to 12 weeks, not a facelift.

Acne

Blue light (which targets acne bacteria) combined with red light (which calms inflammation) shows the clearest benefit in pooled studies (Ngoc et al., 2023). In the UK, NICE recommends considering photodynamic therapy (light plus a light-activated cream, given in clinic) only for adults with moderate to severe acne when other treatments have failed (NICE NG198). It does not recommend home masks. Proven creams and tablets from your GP or pharmacist remain first line.

Hair loss

A 2025 meta-analysis of 38 studies (3,098 patients, mostly androgenetic alopecia, the common pattern hair loss) found hair density increased significantly versus placebo devices after 4 to 26 weeks (Perez et al., 2025). Results varied a lot between studies, and many used FDA-cleared laser caps and helmets rather than face masks. The AAD notes it will not regrow a full head of hair. It is a reasonable add-on to standard treatments such as minoxidil, which the NHS describes, not a replacement.

Pain, joints and healing

The strongest evidence of all is clinical: in cancer care, light therapy halves the risk of painful mouth sores (oral mucositis) caused by chemotherapy and radiotherapy (2026 meta-analysis). For knee osteoarthritis, 22 placebo-controlled trials showed pain fell by about 14 points on a 100-point scale at recommended doses (Stausholm et al., 2019). These results come from professional devices with set doses. A home panel may or may not reach them.

Muscle recovery, sleep and fat loss

Across 23 trials in healthy people, effects on strength and endurance were small and depended on timing and dose (Zhai et al., 2026). Sleep evidence rests on five small trials with wide uncertainty (2026 meta-analysis). For fat loss, three trials (160 people) found about 2.5cm off the waist versus 0.6cm in controls, using specialist lasers (2025 review). None of this supports buying a panel for weight loss.

Thyroid and testosterone

Small studies of light applied over the thyroid in Hashimoto's thyroiditis exist, but the evidence is early and nobody should alter thyroid medication because of a device (2026 review). Testosterone claims rest on cell and animal studies (example), with no good human trials.

Who might reasonably try it

• People with early pattern hair thinning who are already using, or cannot use, standard treatments.
• People wanting a modest improvement in fine lines and skin texture who can commit to near-daily sessions for 2 to 3 months.
• People with knee osteoarthritis pain, ideally guided by a physiotherapist who can advise on dose.

Who should be careful

• Eyes: in 2019 Neutrogena recalled its Light Therapy Acne Mask in the UK and Ireland over a "theoretical risk" of eye injury for people with certain eye conditions or taking medicines that make eyes more light-sensitive (AOP). Use any goggles supplied, and the AAD warns not to swap them for sunglasses.
• Light-sensitising medicines (some antibiotics, acne tablets and others) and light-sensitive conditions such as lupus: check with your GP or pharmacist first (AAD).
• Darker skin tones: visible light can trigger dark patches in some people, so start cautiously.
• Pregnancy: some brands list it as a contraindication because it has not been studied.
• Undiagnosed skin changes: never use light therapy on a changing mole or patch before it has been checked.

If you want to try it: what to look for

1. Stated wavelengths in nm, within 630 to 670 nm (red) and, ideally, 810 to 850 nm (near-infrared).
2. Stated irradiance at a stated distance, so you can calculate the dose.
3. Session guidance that keeps you in the studied range (often a few to around 20 J/cm2 for skin).
4. Eye protection for panels and any device used close to the eyes.
5. A UK seller with a returns policy, UKCA or CE marking and a named company.

If you want a mask, the CurrentBody Skin Series 2 publishes wavelengths and irradiance, and the Omnilux Contour Face uses the two best-studied wavelengths. Both are graded moderate in our mask and panel guide.

Bottom line

Red light therapy is not a fad with nothing behind it, but nor is it the cure-all social media suggests. The best evidence covers pattern hair loss, knee pain, modest skin smoothing and (in hospitals) mouth sores during cancer treatment. Claims about fat loss, sleep, thyroid and testosterone run well ahead of the data. If you buy a device, choose one that tells you its wavelengths and power, protect your eyes, and judge it after 8 to 12 weeks of consistent use.