Blue Light vs Red Light Therapy: What Each One Actually Treats

Affiliate disclosure: This post contains affiliate links. See our affiliate disclosure for details.

Blue light and red light are not two settings on the same treatment. Blue at around 415nm is absorbed by porphyrins inside acne bacteria and destroys them at the skin surface. Red and near-infrared between 630 and 850nm are absorbed by cytochrome c oxidase in your mitochondria and reach into tissue. Different target, different depth, different evidence base. Almost every consumer mask now ships both, so the buying question is not which one works, it is which one you are actually paying for.

We should be straight about where we sit. We own one red light panel, wall-mounted in a spare room, and we have used it most days for over a year. We do not own an LED mask and we have never run a blue-light acne course on ourselves. What follows is the two things we can do honestly: read the trials properly, and pull the published specs off each manufacturer's own page so you can check whether the box can reproduce what the researchers did. On blue light that gap turns out to be the whole story.

Last reviewed: August 2026

Quick comparison

Product	Best For	Price	Coverage	Wavelengths	Irradiance (mW/cm²)	Our rating
Omnilux Clear	Blue light done properly	~$395	Full face, 132 LEDs	415 + 633nm	44 (manufacturer-stated)	4.6
Dr. Dennis Gross FaceWare Pro	Blue and red in one mask	~$455	Full face, 160 LEDs	415 blue, 605 amber, 630 red, 880nm NIR	Not published	4.4
Therabody TheraFace Mask	Most modes, most LEDs	~$650	Full face, 648 LEDs	415 / 633 / 830nm	Not published	4.2
Project E Beauty Lumamask Pro	Budget three-mode mask	~$169	Full face, 72 LEDs	430-445 / 635-644 / 845-855nm	Not published	4.0
LightStim for Acne	One spot at a time	~$169	Handheld, 36 LEDs	415 + 630nm (listing-stated)	Not published	3.8
Solawave Bye Acne	Cheapest FDA-cleared blue	~$69	Handheld spot wand	415 + 630nm	Not published	3.5
Omnilux Contour Face	Red only, full face	~$395	Full face, 132 LEDs	633 + 830nm	~30 (manufacturer-stated)	4.7
Hooga HG300	Red and NIR for the body	~$199	12.2" x 8.2" panel	660 + 850nm	73 at 6"	4.3

The last column is our own rating out of 5, not a retailer star average. Prices and specs were read off each manufacturer's own product page on 29 August 2026 and they move, so confirm the current price before you buy. Two of these carry standing promotions rather than fixed prices: the Lumamask Pro page showed $169 as a 15% discount, and Solawave was running buy-one-get-one-free on the Bye Acne at $69.

What blue light actually does

Blue light works on bacteria, not on you. Cutibacterium acnes (still written Propionibacterium acnes in the older literature) produces porphyrins as a normal part of its metabolism, mostly coproporphyrin. Ashkenazi and colleagues (FEMS Immunology and Medical Microbiology, 2003;35(1):17-24) extracted them, confirmed the emission peak around 612nm under 405nm excitation, and then illuminated live cultures with intense blue light at 407-420nm. The porphyrins absorb, generate reactive oxygen inside the bacterium, and the membrane comes apart. Electron microscopy in that paper shows the structural damage directly.

That is the entire mechanism, and it explains both the strength and the ceiling. Blue light is genuinely antibacterial, and it is antibacterial only where the light lands. Photons at 415nm are scattered and absorbed within the epidermis. They are not reaching a deep nodule, they are not doing anything to your mitochondria, and they have no plausible route to your joints.

The dose numbers in that bench paper are worth sitting with. A single illumination at 75 J/cm² dropped culture viability by under two orders of magnitude. Two consecutive daily illuminations at 75 J/cm² each got four orders, three got five. Now take the only consumer device here that publishes an output figure: the Omnilux Clear at 44 mW/cm². That accumulates 2.64 J/cm² per minute, so 75 J/cm² would take about 28 minutes of continuous exposure, and that 44 figure is the device's total output across 66 red and 66 blue LEDs, not its blue output alone.

We are not saying home masks are useless, and the arithmetic above is not the clinical dose. It is the in-vitro bacterial kill dose, and skin is not a petri dish. But it is the honest scale of the difference between the laboratory numbers people quote in marketing copy and the three-minute session your mask is programmed for.

What the acne trials found

The trial everyone cites is Papageorgiou, Katsambas and Chu (British Journal of Dermatology, 2000;142(5):973-978), and it is worth reading past the headline number. One hundred and seven patients with mild to moderate acne were randomised into four groups: blue light alone (peak 415nm), mixed blue and red light (peaks at 415 and 660nm), cool white light, and 5% benzoyl peroxide cream. Phototherapy groups used portable light sources for 15 minutes daily. Assessment was observer-blinded for the three light arms, though it could not be for the cream.

After 12 weeks the mixed blue-red group showed a mean 76% improvement in inflammatory lesions, with a 95% confidence interval of 66 to 87. That is the number in every product page. What gets left out is the rest of the result. Blue-red beat blue alone at weeks 4 and 8 but not at week 12. It beat benzoyl peroxide at weeks 8 and 12, and it beat white light at every assessment. On comedones, the blackheads and whiteheads rather than the inflamed spots, blue-red managed 58% (CI 45 to 71) and the differences from the other active treatments did not reach significance.

So: a real effect on inflamed lesions, comparable to or better than benzoyl peroxide by week 12, a much weaker case on comedones, and by the end of the study the combination was not measurably ahead of blue on its own.

Zoom out and the picture gets more cautious. The Cochrane review of light therapies for acne (Barbaric et al., Cochrane Database of Systematic Reviews, 2016;9:CD007917) pulled in 71 randomised studies and 4,211 participants. Median study size was 31 people. Interventions varied so much in wavelength, dose and comparator that the authors could not pool the primary outcome at all, and they rated the quality of evidence for participants' own global assessment of improvement as very low. Two thirds of the studies were industry-sponsored. A later systematic review looking specifically at blue light (Diogo et al., Sensors, 2021;21(20):6943) screened 81 articles, included eight, found significant improvements reported across them, and concluded that more detailed trials are needed.

That is a fair summary of where blue light sits in 2026: a mechanism that is well understood and directly demonstrated, clinical results that are positive but built on small and mostly industry-funded studies, and no good answer yet on dose.

What red and near-infrared do instead

Red and near-infrared are absorbed somewhere else entirely. The main chromophore is cytochrome c oxidase, complex IV of the mitochondrial respiratory chain, which contains heme and copper centres that absorb into the red and near-infrared. The leading account (de Freitas and Hamblin, IEEE Journal of Selected Topics in Quantum Electronics, 2016;22(3):7000417) is that photons dissociate inhibitory nitric oxide from the enzyme, electron transport picks up, and mitochondrial membrane potential and ATP production rise. Downstream of that first absorption event you get signalling through reactive oxygen species, cyclic AMP, nitric oxide and calcium, and then changes in gene expression covering protein synthesis, cell migration, anti-inflammatory signalling and antioxidant enzymes.

That is a systemic-sounding chain of events happening in whatever cells the light reaches, which is why red and NIR carry the claims blue never can: collagen, wound healing, joint and muscle pain, hair. It is also why depth matters so much and why the wavelength number on the box is the spec worth checking. We go through that properly in our guide to red light therapy wavelengths, but the short version is that 630nm is a surface wavelength, 660nm is the workhorse red, and 850nm is the one that gets past skin into tissue.

The one thing nobody selling you a blue mask mentions

Blue light darkens some skin.

Duteil and colleagues (Pigment Cell and Melanoma Research, 2014;27(5):822-826) irradiated healthy volunteers with single wavelengths at the two ends of the visible spectrum, 415nm and 630nm, at increasing doses, and compared the result to unexposed and UVB-exposed skin. Colorimetric and clinical assessment showed a clear dose effect with 415nm in both skin type III and type IV subjects. The 630nm red line did not induce hyperpigmentation. And the blue-violet pigmentation was significantly more pronounced than the pigmentation from UVB, lasting up to three months.

This is consistent with what Mahmoud and colleagues (Journal of Investigative Dermatology, 2010;130(8):2092-2097) found using broadband visible light at 400-700nm on 20 volunteers with skin types IV to VI: pigmentation that was darker and more sustained than that from long-wavelength UVA, with no pigmentation observed in skin type II.

Two things follow, and they point in opposite directions, so we will say both. If you have Fitzpatrick I or II skin, this is close to irrelevant to you. If you have type III or darker and you are treating acne, you are using the one wavelength on the mask that is known to trigger dose-dependent pigmentation, on skin that is already prone to post-inflammatory hyperpigmentation from the acne itself. That trade-off deserves a conversation with a dermatologist rather than a decision made off a product page, and we have not seen a single consumer mask listing raise it.

Neither of those studies tested an LED mask. Duteil used a controlled single-wavelength source at defined doses, Mahmoud used a broadband visible source. Nobody has run the home-mask version of that experiment, which means we cannot tell you how much of the effect a four-times-a-week mask protocol reproduces. What we can tell you is which wavelength the signal attaches to, and it is the blue one.

The devices

Omnilux Clear: ~$395, the only one that publishes its output

The Clear runs 415nm blue with 633nm red across 132 LEDs, 66 red and 66 blue, encased in 66 dual-chip bulbs, and Omnilux publishes an output of 44 mW/cm². It is FDA-cleared for acne, and the recommended protocol on the brand's page is four times per week for six weeks. Its blue matches the Papageorgiou trial's 415nm exactly; its red is close to, but not the same as, the trial's 660nm.

That published output is the reason it tops this list. Every other blue device here asks you to take the session length on trust. With 44 mW/cm² you can at least calculate what a session delivers, compare it against the trial protocols, and decide whether the numbers add up. That transparency is rarer than it should be at this price.

Pros: publishes irradiance; the trial-matched 415 plus 633nm pair; FDA-cleared for acne; flexible silicone fits the face properly. Cons: ~$395 for one job; no near-infrared at all; the brand's product page does not state a session length for Clear, which is an odd omission given it publishes everything else.

Best for: anyone with genuine mild to moderate inflammatory acne who wants the version with numbers attached.

Check price →

Dr. Dennis Gross DRx SpectraLite FaceWare Pro: ~$455, both jobs in one shell

Four wavelengths: 415nm blue across 60 LEDs, plus 630nm red, 605nm amber and 880nm near-infrared across 100 more. FDA-cleared, with a pre-programmed three-minute treatment and automatic shut-off.

This is the mask to buy if you want blue and red without owning two devices, and the 880nm inclusion is unusual at this price. What it does not publish is irradiance, so the three-minute session is a number you accept rather than verify. Against the Clear's 44 mW/cm², three minutes would be 7.9 J/cm², and we have no way to know whether this mask is in that neighbourhood or a fraction of it.

Pros: genuinely four wavelengths including 880nm NIR; FDA-cleared; a rigid shell that sits at a consistent distance from the skin. Cons: no published irradiance, which makes the three-minute session unverifiable; rigid masks fit some faces poorly; the most expensive way to get 415nm.

Best for: combination skin where you want acne control and ageing claims from one purchase.

Check price →

Therabody TheraFace Mask: ~$650, 648 LEDs and vibration

Three modes at 415nm blue, 633nm red and 830nm red plus infrared, across 648 LED lights, with cycles between three and nine minutes and added vibration. Therabody's own page markets it as having three times the LEDs of the leading competitor mask.

LED count is the spec we would most like people to stop treating as a proxy for power. It is not one. More LEDs across the same area at lower drive current gets you the same or less delivered energy, and Therabody does not publish an irradiance figure to settle it either way. What you are certainly paying for at $649.99 is the three-mode flexibility and the build, not a demonstrated dose advantage.

Pros: the widest mode range here; long cycles; the 830nm mode is a genuine near-infrared wavelength rather than a deep-red one. Cons: the most expensive device in this comparison by $195; no published irradiance; the LED count claim invites a comparison it cannot back with numbers.

Best for: someone who wants one mask to cover acne, ageing and recovery and is not price-sensitive.

Check price →

Project E Beauty Lumamask Pro: ~$169, the cheap way to get all three

Published as ranges rather than single figures, which we appreciate: red at 635-644nm, blue at 430-445nm, infrared at 845-855nm, across 72 LEDs in 36 bulbs, with a 12-minute auto shut-off. FDA-cleared.

Note the blue band. At 430-445nm this sits above the 415nm used in the acne trials and above the 407-420nm window in the porphyrin work. Porphyrin absorption falls away sharply past the Soret peak near 405-415nm, so this is not the same blue the evidence was built on. The 12-minute session is the longest here by a distance, which may partly compensate, but no irradiance figure is published so that stays a guess.

Pros: three wavelength bands at a quarter of the TheraFace price; publishes ranges rather than a marketing single figure; the longest session length here. Cons: blue band centred well above 415nm; no published irradiance; 72 LEDs is the lowest count in the mask group.

Best for: trying LED light therapy without committing $400.

Check price →

LightStim for Acne: ~$169, treat one spot properly

A handheld with 36 LEDs, FDA-cleared to treat mild to moderate acne, used no more than three minutes per area once daily. The brand's own product page does not publish wavelengths at all. LightStim's Amazon listing states 415 and 630nm, and since that listing is the brand's, we will use it, but a device sold on its FDA clearance ought to put its wavelengths on its own product page.

The handheld format is a real choice, not a compromise. If you get two or three breakouts a month rather than widespread acne, three minutes on each is a more sensible use of light than illuminating your whole face four times a week.

Pros: FDA-cleared for the specific indication; a sensible format for occasional spots; a fifth of the cost of a comparable full-face mask. Cons: wavelengths absent from the brand's own page; no published irradiance; treating a full face one area at a time is slow.

Best for: occasional breakouts rather than a standing acne condition.

Check price →

Solawave Bye Acne: ~$69, the cheapest FDA-cleared blue

415nm blue with 630nm red, three minutes per area, FDA-cleared, USB-C rechargeable. At $69 on Solawave's own page (running buy-one-get-one-free when we checked on 29 August 2026) it is the cheapest legitimate route to 415nm we found.

No irradiance figure, so the same caveat as the rest applies, and a spot wand cannot do what a mask does. But it has the trial-matched wavelength pair and a clearance, at a price where getting it wrong costs you a takeaway rather than a month's rent.

Pros: the correct 415 plus 630nm pair at the lowest price here; FDA-cleared; genuinely portable. Cons: no published irradiance; spot coverage only; the standing BOGO makes the real per-unit price hard to pin down.

Best for: testing whether blue light does anything for your skin before spending $400 to find out.

Check price →

Omnilux Contour Face: ~$395, what you buy if your skin is clear

The Clear's sibling, and the honest recommendation for most people reading this. 633nm red with 830nm near-infrared across 132 LEDs, 66 of each, at a published output of about 30 mW/cm². No blue light at all. The recommended protocol is three to five times per week for four to six weeks, then maintenance.

If your acne is not active, blue is a mode you will switch on twice and never touch again. Buying the red and NIR version means the whole 132-LED array is doing the job you actually want, rather than half of it. We cover the wider mask field in our best red light therapy masks roundup.

Pros: publishes irradiance; every LED serves the job most buyers want; the 830nm is real near-infrared. Cons: ~30 mW/cm² is below the Clear's 44; no acne indication; same price as the Clear, so the choice is genuinely either/or.

Best for: collagen, fine lines and general skin work with no active acne.

Check price →

Hooga HG300: ~$199, when the target is not your face

The reference point for what red and NIR do that blue cannot. Sixty 5W LEDs at 660nm and 850nm in a 12.2 by 8.2 by 2.5 inch panel drawing 80W at the wall, with published irradiance at four distances: 100 mW/cm² at 3 inches, 73 at 6, 41 at 12 and 24 at 18.

Publishing the distance alongside the number is the single most useful thing a panel brand does, and most of them still do not. It is also the spec that makes clear why a mask and a panel are not comparable devices: a mask sits on your skin at low output over a small area, a panel throws far more energy across a shoulder or a knee. If panels are where you are heading, start with our best red light therapy panels comparison.

Pros: irradiance published at four distances; independent red and NIR controls; the cheapest device here per watt delivered. Cons: no blue and no facial fit; a fan that is audible in a quiet room; you have to hold a position in front of it.

Best for: muscle, joint and recovery work, where blue light has no role at all.

Check price →

Buyer's guide

Start with whether you have active acne. That single question decides this. Blue light has one job. If you do not currently have inflammatory breakouts, every dollar of the blue half of a dual-mode device is dead weight, and the red-only version of the same mask is usually the same price with twice the useful LEDs.

Check that the blue is actually 415nm. The porphyrin absorption peak sits near 405-415nm and falls off quickly above it. A mask advertising "blue light" at 460nm is not delivering what the acne trials delivered. One of the six blue-capable devices here, the Lumamask Pro, publishes a blue band of 430-445nm, and nothing in its marketing copy flags the difference.

Treat irradiance as the spec that separates the field. Only three of these eight publish one. Without it, a session length is a number the brand chose, and you have no way to compare a three-minute cycle on one mask against a twelve-minute cycle on another. We are not saying an unpublished figure means a bad device. We are saying that among devices you cannot otherwise tell apart, the one showing its working is the safer bet.

Know what FDA-cleared means here. All six masks and wands on this list say it, and it is not a claim about how well the device performs. It means the FDA accepted the device as substantially equivalent to something already on the market for that indication. Our guide to FDA-cleared red light therapy devices works through the difference between cleared, approved and registered, which brands blur constantly.

If your skin is type III or darker, weigh the pigmentation evidence. 415nm produced dose-dependent hyperpigmentation in type III and IV volunteers in the Duteil study, more pronounced and longer-lasting than the pigmentation from UVB, while 630nm produced none. That is a controlled-source finding rather than a mask finding, and it is still the strongest signal available on the question.

FAQ

What is the difference between blue light and red light therapy?

Blue light at around 415nm is absorbed by porphyrins inside acne bacteria and kills them at the skin surface. Red and near-infrared light between 630 and 850nm is absorbed by cytochrome c oxidase in your mitochondria, raising ATP production and triggering anti-inflammatory and repair signalling in the tissue the light reaches. Blue is antibacterial and shallow, red is metabolic and deeper. That is why blue is only ever sold for acne, while red and near-infrared carry claims for collagen, wound healing, pain and hair.

Is blue light or red light better for acne?

The best-known trial found a mean 76% improvement in inflammatory lesions after 12 weeks of daily 15-minute mixed blue and red phototherapy (Papageorgiou et al., 2000), better than 5% benzoyl peroxide at weeks 8 and 12. The same study found the combination was significantly better than blue alone at weeks 4 and 8, but not at week 12. So the combination works faster, and by three months blue on its own had largely caught up. On comedones rather than inflamed spots, neither did much.

Does blue light therapy really kill acne bacteria?

Yes, and the mechanism is directly demonstrated rather than inferred. Cutibacterium acnes produces coproporphyrin, which absorbs blue light around 407-420nm and generates reactive oxygen inside the bacterium. Ashkenazi et al. (2003) showed viability dropping by up to five orders of magnitude after repeated illumination, with membrane damage visible under electron microscopy. The open question is dose: the bench work used 75 J/cm² per illumination, and no consumer device publishes enough information for you to know what fraction of that a three-minute session delivers.

Can blue light therapy darken your skin?

It can, and this is the least-discussed risk in the category. Duteil et al. (2014) exposed volunteers to 415nm and 630nm at increasing doses. The 415nm blue-violet light produced dose-dependent hyperpigmentation in both skin type III and type IV subjects, significantly more pronounced than the pigmentation from UVB and lasting up to three months. The 630nm red light produced none. Those were controlled single-wavelength sources rather than LED masks, so nobody has measured how much a home protocol reproduces, but if you have type III or darker skin it is worth raising with a dermatologist first.

Do I need a mask with both red and blue light?

Only if you have active inflammatory acne. If you do not, the blue mode is capacity you are paying for and never using, and a red-only mask at the same price gives you twice as many LEDs doing the job you want. Omnilux sells this choice cleanly: the Clear is 66 red plus 66 blue for acne, the Contour Face is 66 red plus 66 near-infrared for skin, both at ~$395. If your goal is fine lines rather than breakouts, buy the second one.

What wavelength is blue light therapy?

The wavelength with the evidence behind it is 415nm, which sits close to the Soret absorption peak of bacterial porphyrins. The acne trials used 415nm, and the bench work on bacterial killing used 407-420nm. Porphyrin absorption drops off sharply above that window, so a device advertising blue light at 445 or 460nm is not delivering the same thing. Check the number, not the colour name, because "blue light" on a product page can mean anything from 405 to 470nm.

How long does blue light therapy take to work on acne?

The trial that established the effect ran daily 15-minute sessions and assessed every four weeks, with the improvement in inflammatory lesions still building at week 12. Omnilux's protocol for the Clear is four sessions a week for six weeks. Expect to judge it at the six-week mark rather than the second week, and expect it to work better on inflamed spots than on blackheads, where the same trial found a weaker and non-significant result.

Is red light therapy or blue light therapy better for wrinkles?

Red, without qualification. Blue light has no mechanism for collagen: it is absorbed at the skin surface by bacterial porphyrins, not by anything in your dermis. The collagen claims come from red and near-infrared photobiomodulation acting on mitochondria in fibroblasts. If wrinkles are your goal, a red plus near-infrared device is the correct purchase and the blue mode is irrelevant. Our guide to red light therapy for skin covers what the evidence supports there.

Our Verdict

If we were starting again and had active acne, we would buy the Omnilux Clear at ~$395. It is the only device here that publishes both its wavelengths and its output, its blue matches the Papageorgiou trial's 415nm exactly, and it is FDA-cleared for the indication. Paying more for the TheraFace gets you more modes and no more information.

But most people reading a blue versus red comparison do not have active acne, and for them the honest answer is that blue light is not the purchase. The Omnilux Contour Face, same price, same 132 LEDs, all of them red and near-infrared, is the better buy. If the target is a shoulder or a knee rather than a face, neither mask is the right tool and the Hooga HG300 at ~$199 does more delivered work than any of them.

The thing we would most like to change about this category is the missing irradiance figures. Five of these eight devices ask you to trust a session length without publishing the output that makes it meaningful. Until that changes, the brands that show their working deserve the benefit of the doubt, and the ones that do not deserve rather less of it.

More on this: our red light therapy hub, and about BankrollZen.