Short answer: The real red light therapy dangers are narrow and specific: direct eye exposure to high-irradiance devices, overdosing on the assumption that more light is better, pigment reactions in darker skin, photosensitising medication, and thermal injury from poorly built panels. Everything else attributed to red light is either device-quality failure or marketing invention.

Why a risk page, and how to read it

We install professional photobiomodulation equipment in clinics, spas and private homes, so we have a commercial interest in you buying one. We would still rather write down the six things that actually go wrong than deal with them after commissioning.

Two things get called red light therapy dangers that are not. First, “it’s radiation” — it is non-ionising visible and near-infrared light, in the same band as a sunset, with no ultraviolet component in a properly specified device. Second, “EMF”. Panels do emit low-frequency fields from their drivers, at levels comparable with any mains appliance, and no evidence connects that to harm. Chasing EMF numbers distracts from the risks below, which are real.

If you want the efficacy side rather than the risk side, start with our complete red light therapy guide. This page assumes you already know what the therapy is and want to know what can go wrong.

Risk 1: Eye exposure — the one that deserves genuine caution

This is first because it is the only red light therapy danger capable of permanent injury, and the only one where the protective behaviour is not intuitive.

The eye’s defences are tuned to visible light. Bright red triggers a blink and an aversion response. Near-infrared at 810–850 nm, which every serious panel and bed emits, is largely invisible — so a device can deliver substantial irradiance to the retina while feeling comfortable. There is no reflex to protect you from light you cannot see.

What this means in practice:

• Never look directly into a panel or bed array, at any distance, for any duration. The dose reaching the retina scales with irradiance and time exactly as it does at the skin.
• Use the goggles supplied with the device on anything above face-mask power. Closed eyelids attenuate red light but transmit near-infrared far more readily.
• A face mask is a special case. Masks sit centimetres from the eye. Buy one with opaque eye coverage or integrated shields, not one that relies on you keeping your eyes shut.
• Staff exposure is the forgotten one. A therapist working an eight-hour shift in a treatment room accumulates far more exposure than any client. Treatment rooms need the operator outside the beam or behind screening.

Worth saying clearly, because the internet gets this wrong in both directions: the “blue light hazard” is a real laboratory phenomenon under abnormally intense exposure, but a 2022 perspective in the American Journal of Ophthalmology argued it has been commercially misused to alarm people about ambient light, with no evidence that ordinary exposure causes macular degeneration (Mainster et al., 2022). Red and near-infrared are not blue light. Protect your eyes from a red light bed because of its irradiance, not its colour.

Our red light therapy clinic page covers how the room itself should be laid out to keep both client and operator out of stray beam.

Risk 2: Dose — more light is not more benefit

This is the most consequential misunderstanding in the category, and the reason many people conclude the therapy “does nothing” when they have simply overshot.

Photobiomodulation follows a biphasic dose-response, sometimes called the Arndt-Schulz curve: below a threshold nothing happens, in a window the effect is positive, and above that window the effect diminishes and can invert. This is not fringe theory — it is the framework the field itself uses, laid out by Huang and colleagues at Harvard and Massachusetts General Hospital (Huang et al., 2009) and updated two years later with in vitro and animal evidence for biphasic patterns in ATP production and mitochondrial membrane potential (Huang et al., 2011).

What this means for a user:

• Doubling the session time does not double the result. It may reduce it.
• Standing closer is a dose change. Irradiance falls off steeply with distance, so halving the distance can multiply the delivered dose several times over. The manufacturer’s stated distance is part of the prescription.
• Stacking devices compounds the error. A mask, then a panel, then a bed, in one evening, is not a protocol.
• Devices that publish no irradiance figure cannot be dosed at all. Without mW/cm² at a stated distance, there is no way to calculate J/cm², which means no way to know where you are on the curve.

The practical harm of overdosing is usually wasted time rather than injury — but it is why “red light therapy didn’t work for me” is such a common verdict. Our red light therapy at home guide covers what to check on a spec sheet.

Risk 3: Skin type and pigment reactions

Red light is generally well tolerated across skin types, so this is a caution rather than a prohibition. It is still real, and almost never mentioned on device packaging.

Visible light — not only ultraviolet — induces melanogenesis. An ex vivo study of skin from 22 patients with facial melasma found effective melanogenesis in the basal layer after exposure to visible light as well as UVB and UVA, in both melasma-affected and adjacent normal skin (Alcantara et al., 2020). The response is more pronounced in Fitzpatrick phototypes IV to VI.

So the honest guidance is:

• If you have melasma or post-inflammatory hyperpigmentation, treat facial red light as something to discuss with a dermatologist first, and watch the treated area for the first few weeks. Some people see improvement; some see the patch darken.
• Darker skin types should start at shorter sessions and increase only if there is no pigment change, rather than starting at the maximum the manual allows.
• Do not use red light over a fresh laser, peel or microneedling site without clearance from whoever performed it.

The dermatological upside is well documented — a widely cited review covers collagen production, wound healing and skin restoration under red and near-infrared wavelengths (Avci et al., 2013). The point is that pigment is a variable in that response, not a fixed constant.

Risk 4: Photosensitising medication

This is the risk most likely to catch someone out, because the person taking the drug rarely connects it to a light device.

Drug-induced photosensitivity is among the most common cutaneous adverse drug reactions, and it is triggered by ultraviolet or visible radiation. Several hundred drugs are implicated, including common non-steroidal anti-inflammatories, cardiovascular drugs, psychotropics, antimicrobials, lipid-lowering agents and antineoplastics, with reactions ranging from erythema and swelling through blistering to lasting hyperpigmentation (Kowalska et al., 2021).

The practical rules are simple:

• Any new client or family member on regular medication should check the photosensitivity warning on their leaflet before a first session.
• Topical retinoids, St John’s wort, doxycycline, amiodarone and several diuretics are the ones we see most often.
• Commercial facilities should have this on the intake form, not in a verbal question at the door.

Risk 5: Thermal injury and build quality

Red and near-infrared light is not a heat therapy — a properly specified session leaves you at normal body temperature. But the devices get hot, and cheap ones get hot in the wrong places.

The failure modes we have seen in the field:

• Hot diode housings on contact devices. Wands and masks pressed against skin for twenty minutes can cause low-temperature contact burns if the housing has no thermal management.
• Driver failure in enclosed panels. Panels mounted flush to a wall with no rear air gap run their drivers hot, which shortens life and, in the worst cases, ends in a burnt connector.
• Prolonged low-grade heat over the same patch of skin is the mechanism behind erythema ab igne, the reticulated pigmentation seen from laptops and heat packs — rare with red light, but not impossible with a badly built contact device used daily.

None of this is inherent to photobiomodulation. It is inherent to buying a device with an unbranded driver and no thermal design. This is exactly the line we draw in our red light sauna and bed guide between consumer gadgets and professional equipment.

Risk 6: Unverifiable specifications

The most common danger in this market is not physiological at all. It is that you cannot tell what you bought.

Irradiance inflation is endemic. “200 mW/cm²” means nothing without the measurement distance, and many listings quote the value at the diode surface — a number no user ever experiences. Wavelength claims are approximate, diode counts are quoted instead of coverage, and “FDA cleared” may refer to a general wellness classification rather than any clinical claim.

The four questions that separate a real device from a lit-up plastic box:

1. What is the irradiance in mW/cm², at what distance, measured by whom?
2. What are the exact peak wavelengths, and what is the spectral spread around them?
3. What regulatory clearance or CE marking applies, and to what claim?
4. What session time and distance does the manufacturer specify, and does that give a dose inside the published window?

A supplier who cannot answer all four has not tested the product. Compare device classes in our red light therapy review.

Contraindications, and how firm each one is

Situation	Position	Why
Active cancer in the treatment field	Physician decision	A 2020 systematic review of 67 studies found current human data, supported by most animal work, shows photobiomodulation is safe at recommended clinical parameters — but this is an oncologist’s call, not a spa’s
Pregnancy	Avoid over the abdomen	No evidence of harm; no evidence of safety either. Precaution, not alarm
Photosensitising medication	Clear it first	Documented reaction pathway with visible light
Photosensitivity disorders (lupus, porphyria)	Physician decision	The trigger is light, not wavelength band
Epilepsy	Avoid pulsed modes	Continuous-wave output is not a flicker stimulus; pulsed devices can be
Thyroid, over the neck	Discuss first	Commonly repeated caution with thin evidence either way
Fresh injectables, peels or laser	Wait for clearance	Practitioner’s call on healing timeline

The oncology position deserves a note because it is the one most often stated wrongly in both directions. The systematic review found no evidence that photobiomodulation at clinical parameters promotes tumour growth or recurrence, and clinical benefit in preventing cancer-therapy complications (Bensadoun et al., 2020). That is reassuring, and it is still not a reason for a wellness facility to treat an oncology patient without their consultant’s agreement.

Four supposed red light therapy dangers that are not real

• “It causes skin cancer.” No. Red and near-infrared are non-ionising and non-UV. A device emitting UV is a different device, badly built or mislabelled.
• “It damages eyes at any dose.” Overstated. A face mask at a few mW/cm² with the eyes covered is not the same exposure as staring into a bed array. The risk scales with irradiance.
• “EMF from panels is dangerous.” No evidence supports this. Buy on measured irradiance and build quality, not on EMF marketing.
• “It detoxifies.” Not a danger, just not a thing. The mechanism is mitochondrial light absorption, and there is no elimination pathway involved.

What a safe installation looks like

When we specify a red light room — whether inside a longevity centre, a commercial spa or a private residence — the safety measures are physical, not procedural. Signs get ignored; geometry does not.

• Fixed distance markers on the floor or a bench at the manufacturer’s stated treatment distance, so the dose is not left to guesswork.
• A hard session timer that ends the session, rather than a suggestion in a manual.
• Goggles stored at the entry point, in the path the user walks, not in a drawer.
• Operator position outside the beam, or a screen, for any staffed treatment.
• Rear ventilation clearance behind panels and a dedicated circuit for beds.
• Wipeable, non-reflective surfaces. Gloss and mirrors put stray light where nobody accounted for it.

The same discipline applies to combined cabins, where light sits alongside heat: our infrared sauna guide explains why the two are different stimuli, and the infrared sauna range covers the build. The same specification logic applies to PEMF and every other modality on the floor — see building a recovery and longevity space, wellness centre design, or private spa for a residence.

Frequently asked questions

What are the real dangers of red light therapy?

Direct eye exposure to high-irradiance panels and beds is the only one capable of permanent injury, because near-infrared is invisible and triggers no blink reflex. After that: overdosing past the biphasic window, pigment reactions in Fitzpatrick IV–VI skin and melasma, reactions in people on photosensitising medication, and contact burns from badly built devices. Red light is non-ionising and non-UV, so it does not cause skin cancer.

Do you need goggles for red light therapy?

For anything above face-mask power, yes. Near-infrared at 810–850 nm is largely invisible, so the eye has no aversion response to it, and closed eyelids transmit it far more readily than they do red light. Use the goggles supplied with the device, and keep operators out of the beam in staffed rooms.

Can you overdo red light therapy?

Yes. Photobiomodulation follows a biphasic dose-response: below a threshold nothing happens, within a window the effect is positive, and above it the benefit diminishes and can reverse. Longer sessions and standing closer both increase dose. If a device publishes no irradiance figure at a stated distance, the dose cannot be calculated at all.

Who should not use red light therapy?

Anyone on photosensitising medication until they have checked with a pharmacist or doctor; anyone with a photosensitivity disorder such as lupus or porphyria; anyone with active cancer in the treatment field without their oncologist’s agreement. Pregnancy is a precaution over the abdomen rather than a documented risk, and pulsed devices should be avoided in epilepsy.

Specifying it properly

Sauna Dekor has been manufacturing wellness installations since 1987 — 40 years of saunas, hammams, steam rooms and, more recently, the red light rooms and cabins that sit beside them. Most of the risks on this page are solved at specification stage: a device with published, measured irradiance, a room laid out so the beam goes where you intended, and a timer that ends the session. Talk to us before the equipment is ordered.