Category: Beauty & Skincare

Buyer’s guide

Some links in this post are affiliate links. If you buy through them I earn a commission at no extra cost to you. I bought the mask I own with my own money before any brand relationship existed. I am not a doctor and nothing here is medical advice.

Written by Anna Katherine Page, who spent eighteen months choosing an LED mask and then two years finding out what the spec sheet actually meant. Last updated September 4, 2026.

Why is it so hard to compare LED masks?

Because the numbers on the box are chosen to be impressive rather than comparable, and the one number that would let you compare devices is the one most brands do not publish.

I spent a year and a half choosing my first mask. Here is what I said on camera when I finally bought one:

“I even went into ChatGPT and compared a couple of the very top masks on the market, including the Dr Dennis Gross and the Omnilux, to this one. It has a higher number of LED lights than either of those masks.”

That was true, and it was also the wrong thing to be comparing. I want to walk you through what I would check now.

1. Do not compare LED counts

Modern LED masks use multi-chip packages. One physical emitter on the inside of the mask contains three or four separate light-emitting dies, often of different colours. A brand can honestly describe that as four LEDs.

So the mask I own is advertised as 272 LEDs on Amazon and 276 in one of the brand’s own images, while the brand’s own FAQ says “69 pieces.” An independent spectrometer review counted 68 quad-chip packages. All of those numbers describe the same mask.

Once you know this, the arms race reads differently. A “480 light source” mask with 160 tri-wick emitters and a “432 LED” mask with 108 quad-wick emitters are not obviously different products.

2. Look for irradiance, and notice who will not tell you

Irradiance is the power density hitting your skin, in milliwatts per square centimetre. Multiply it by seconds and you get fluence, the dose, in joules per square centimetre. That is the number every study in this field reports and it is the number that determines whether a device is delivering anything.

Here is what the major brands publish.

One brand out of six publishes a real figure. That is the finding, and it is not unique to LED masks: when I compared at-home IPL hair removal devices, not one competing brand published an energy figure anywhere on its own website, and neither did the clinic platforms costing tens of thousands.

The mask I own is the odd one out for publishing a number at all, and it published one that does not survive checking. Its listing claims 105 mW/cm². Independent measurement put it near 30. The FDA filing from the factory that builds INIA masks lists 3.5 to 27 mW/cm² depending on model and mode.

So my practical advice is uncomfortable but honest: a published irradiance figure is only useful if it is true, and you usually cannot tell. What you can do is treat a suspiciously large number as a warning rather than a feature. Real cleared consumer masks measure around 30 mW/cm². A claim of 100-plus is either a different measurement method or marketing.

The useful half of that story is what the measurement found rather than what the box claimed. At roughly 30 mW/cm², the mask I bought for about $130 is delivering the same measured output that Omnilux publishes for a mask at $395. That is the thing worth knowing when you are standing in front of a wall of options wondering whether the price gap buys you anything.

3. Ask for the K-number

This is the check almost nobody does and it takes sixty seconds.

Consumer LED face masks are Class II devices that require FDA 510(k) premarket notification. There are two relevant product codes: OHS for over-the-counter light-based wrinkle reduction, and OLP for over-the-counter light-based acne treatment. There are roughly 162 clearances under OHS and 51 under OLP.

A real clearance has a K-number, like K250966. You can search any brand or manufacturer name in FDA’s 510(k) database and see for yourself.

Three things to know before you do:

“FDA registered” is not clearance. FDA’s own consumer page is unusually blunt about this. On phrases like “FDA registered” and “FDA certified” it says: “Such words may be used to mislead you. Is that the same thing as FDA approved, FDA cleared, or FDA authorized? The short answer is NO.” Registration means a facility paid a fee and told FDA it exists. FDA does not issue registration certificates, and companies using fake ones “misbrand the device.”

Clearance is not a finding that it works. A 510(k) demonstrates substantial equivalence to a device already on the market. The pathway that requires evidence of safety and effectiveness is PMA, and no consumer LED mask goes through it. Nothing in this category is “FDA approved.”

One clearance can cover many masks. This is the detail that surprised me most. Clearance K243040 covers nine model numbers, eight of them the factory’s house codes and one a customer’s. The irradiance table in that single filing spans 3.5 to 27 mW/cm² across the models it covers. A brand can accurately say “covered by a 510(k)” while shipping the weakest variant in the family.

4. Understand that you are probably buying a white-label product

This is not a scandal. It is how the category works, and knowing it will save you money.

Most LED masks are designed and built by a small number of contract manufacturers in Guangdong province and sold under many brand names. The clearance for the CurrentBody Skin LED Mask Series 2, K250966, is held by Shenzhen Kaiyan Medical Equipment, not by CurrentBody. The mask I own is linked in FDA’s device identifier database to Guangdong Newdermo Biotech, and its current models are listed by a different manufacturer again, Shenzhen Siken 3D.

The practical consequence: a large part of what separates a $450 mask from a $130 mask is the brand, not the hardware. Not all of it. Fit, build quality, silicone grade, timer behaviour, battery engineering and customer support are real and they do differ. But the emitters and the wavelengths often come from the same supply chain.

This is exactly why the irradiance question matters more than the price. When my $130 mask measures the same 30 mW/cm² that Omnilux publishes for its $395 mask, that is the white-label reality showing through.

5. Check wireless, timer, and eye protection

Wireless is the feature that decides whether you keep using it. I will die on this hill. A corded mask means you have to be near an outlet, sitting still, for ten minutes, three times a week, forever. A wireless mask with a swappable battery means you can do it while doing something else. Two years in, that is the single reason I still use mine, and adherence is the only variable in this whole category that is fully in your control.

A built-in timer is a dose control, not a convenience. Photobiomodulation has a biphasic dose response, meaning past an optimal dose the effect diminishes and can reverse. Ten minutes at 30 mW/cm² delivers about 18 J/cm², at the top of the standard stimulation range. A mask that shuts itself off is protecting you from yourself. I go into this in how often you should actually use red light therapy on your face.

Eye protection should be included and you should use it. FDA’s adverse event database holds 100 reports under the OHS product code, and 25 of them mention the eyes, including reports of blurred vision, severe eye irritation and light sensitivity. These are unverified passive reports and causality is not established, but the pattern is worth knowing. Johnson & Johnson recalled the Neutrogena Light Therapy Acne Mask in July 2019 over risk of eye injury. Some brands claim clearance means goggles are unnecessary. Clearance is not an eye-safety certification. The relevant standard is IEC 62471, and very few brands claim conformance to it.

What should you actually buy?

Work backwards from your concern.

General texture and tone: red around 630 to 660nm plus near-infrared at 830 to 850nm. This is the mainstream combination and where such evidence as exists sits. Any mask with those two wavelengths, a ten minute timer, and eye pads will do the job as well as any other.

Acne: you want true blue at 415 to 460nm, and you should know the evidence is not what the marketing suggests. The meta-analysis of 14 randomised trials and 698 participants found no significant effect on either inflammatory or non-inflammatory lesion counts. Blue light’s real advantage in the data is tolerability, not superiority. In one head-to-head, benzoyl peroxide outperformed it on lesion reduction, 31 percent against 22, with side effects in 93 percent of the benzoyl peroxide group against 23 percent for light.

Dark spots: do not buy an LED mask. There is no controlled evidence and there is a real argument it is the wrong direction. I wrote that up in full in does red light therapy work for dark spots.

Wrinkles specifically: know that the best-designed sham-controlled facial trial, 95 women, double-blind, found no significant difference from sham on its blinded clinician-rated outcome. A retinoid has a much stronger evidence base for the same problem. Buy the mask as an addition, not a replacement.

The sixty-second checklist

Frequently asked questions

Related reading: my two-year INIA mask review · how often to use red light therapy on your face · does red light therapy work for dark spots · the same spec problem in at-home IPL

What the research says

Some links in this post are affiliate links. If you buy through them I earn a commission at no extra cost to you. I bought the LED mask I own with my own money, partly for this exact reason, and it did not do what I hoped. I am not a doctor and nothing here is medical advice.

Written by Anna Katherine Page, who bought an LED mask for hyperpigmentation and went looking for the evidence afterwards. Last updated September 4, 2026.

Why I went looking for this

I bought my LED mask in late 2024 after a year and a half of research, and I said this on camera at the time:

“I have a lot of concerns with hyperpigmentation in my skin, and also acne.”

Those were my two reasons. Two years later, having gone through the actual literature for this post, I have to tell you that I picked a device for two jobs it has almost no evidence of doing. I still use the mask, and I still think it is a reasonable purchase for other reasons, which I go through in my full review. But I was wrong about why I was buying it, and if you are about to make the same purchase for the same reason, I would rather you heard it here.

Is there any evidence red light therapy fades dark spots?

I searched for randomised controlled trials of red LED for hyperpigmentation, dark spots and post-inflammatory hyperpigmentation. Every result on the first pages was a device retailer or a content farm. No primary trial surfaced.

The most useful negative evidence is a 2025 narrative review in Bratislava Medical Journal covering 59 red light studies and 1,882 patients. It enumerates the dermatological indications where red light has been studied: acne, basal cell carcinoma, squamous cell carcinoma, actinic keratosis, rosacea, psoriasis, hidradenitis suppurativa, mycosis fungoides, warts, vitiligo, necrobiosis lipoidica, granuloma annulare and photoaging.

Pigmentary disorders are not on that list. Not as a weak indication. Not at all.

What happened when someone actually tested LED on melasma?

Somebody did run the trial, and the result is worth knowing about.

Galache and colleagues, publishing in Lasers in Medical Science in 2025, ran a randomised, double-blind non-inferiority trial in 21 women with melasma over twelve weeks. One group got amber LED photobiomodulation plus a placebo cream. The other got a sham LED device plus 5 percent liposomal tranexamic acid cream, which is a real melasma treatment.

The finding: “No significant differences were observed in the MASI or PGA scores, within or between groups.” MASI is the standard melasma severity index. Neither treatment moved it.

But quality of life scores improved in both arms. That combination, a subjective measure improving while the objective measure sits still, in both the active and the control group, is close to a textbook picture of placebo response.

The trial was small and underpowered at 51 percent and terminated early, so it does not prove LED does nothing. What it does is remove the excuse that nobody has looked.

Can red light therapy make pigmentation worse?

This is the part that changed how I think about the whole category.

Visible light causes pigmentation. This is well established and it is why visible-light-blocking sunscreens exist. A 2026 review in Dermatology and Therapy puts the numbers on it: blue light “is the most potent inducer of pigmentation at fluences of 40 to 80 J/cm², while red light requires much higher fluences, approximately 150 J/cm².”

The same review states that visible-light-induced pigmentation “is more sustained, persisting 10 to 22 days, is deeper in higher skin phototypes, and increases with increased exposure,” and that “conditions such as melasma and post-inflammatory hyperpigmentation are exacerbated by” visible light. The mechanism is Opsin-3, a receptor in melanocytes that responds to blue light by driving melanin production.

Now do the arithmetic on a mask.

Red light in a single mask session sits comfortably below the threshold. That is reassuring and I want to say it clearly.

Blue is a different conversation. Its threshold is roughly half of red’s, and the acne protocols that produced results in the literature ran daily for twelve weeks. Papageorgiou’s landmark blue and red acne trial accumulated around 320 J/cm².

There is also real-world signal. FDA’s adverse event database holds a report from July 2026 describing “burn marks or hyperpigmentation” following prolonged use of an LED face mask, and another describing “dark discoloration and bruising” on a user’s face after using a different mask. Those are unverified passive reports and causality is not established, which is exactly how they should be read. They are not nothing either.

And a 2026 review specifically records a lupus-like reaction following red LED mask therapy, while a melasma-focused review states that “recent professional opinions warn against the uncontrolled application of red and blue LED mask devices in melasma-prone individuals, as the stimulation of melanocytes with a subsequent increase of heat is possible.”

Does skin tone change the answer?

Yes, in two directions, and both matter.

On safety: light-induced pigmentation is deeper and longer-lasting in Fitzpatrick IV to VI. If you have a deeper skin tone and a pigmentary concern, an unsupervised home light device is a worse bet for you than for someone with fair skin, not a better one.

On effectiveness: melanin absorbs light that was aimed at something else. The 2025 review notes a differential efficacy signal, with one study finding a lower clearance rate in individuals with pigmented skin even after accounting for other lesion characteristics. Equal device output is not equal delivered dose across skin types, and no facial trial has adjusted dose for this.

The claim you will see everywhere is “safe for all skin types.” What the literature actually supports is narrower: LED photobiomodulation is non-ablative and athermal by design, so it does not carry the burn and post-inflammatory hyperpigmentation risk that IPL and ablative lasers do. That is a real and defensible advantage. It is not the same as a documented finding of safety across all phototypes, and the reviews themselves describe skin-type reporting in this literature as limited.

This is one place where the at-home device genuinely does beat the clinic. I have written about the other side of it, where going to a clinic does not solve the skin tone limit either, in my comparison of at-home IPL against an in-office photofacial.

What actually works on dark spots?

I am not going to leave you with only a no.

Sunscreen that blocks visible light. This is the single highest-return move and the most skipped. Because visible light drives pigmentation, a standard UV-only sunscreen is not enough for melasma. Trials of sunscreens combining UV protection with visible-light protection have shown benefit in preventing melasma relapse. Tinted formulas containing iron oxides are the practical version of this.

Topical tyrosinase inhibitors. Tranexamic acid, azelaic acid, and the prescription options your dermatologist will discuss. These act on the pigment pathway itself, which is the part an LED mask never touches.

Retinoids. Topical tretinoin for photodamage has a systematic review and meta-analysis of randomised trials behind it, which is more than anything in the LED literature can say. I use the PRMR Youth Radiance serum.

In-office IPL, with caveats. A 2025 systematic review pooling 41 trials and 3,234 participants found IPL success rates of 74.6 to 90 percent on solar lentigines, comparable to picosecond lasers. That is a real, large evidence base. It comes with genuine risk in deeper skin tones and it is a clinic procedure, not a device you buy.

What to do if you have melasma specifically: see a dermatologist rather than buying a device. Melasma is the one pigmentary condition where the wrong light treatment can set you back months.

So should you buy an LED mask at all?

Maybe, but not for this.

The honest case for a home LED mask is red and near-infrared light on general skin texture and tone, where the evidence is modest and mixed but not absent. Buy it knowing the ceiling is low, the timeline is six weeks minimum, and the best-designed sham-controlled facial trial found nothing on its blinded primary outcome.

Do not buy it as a dark spot treatment. That purchase has no evidence behind it, and if you have melasma it may be actively the wrong move.

The mask I own and use is the INIA GLOW. It measures about 30 mW/cm² independently, the same as a mask that costs three times more, and it is wireless, which is the reason I have kept using it for two years. I bought it for my hyperpigmentation. I keep it for other reasons.

Frequently asked questions

Related reading: my two-year INIA mask review · how often to use red light therapy on your face · at-home IPL versus an in-office photofacial · the retinol serum I use

How to use it

Some links in this post are affiliate links. If you buy through them I earn a commission at no extra cost to you. I bought the mask I use with my own money. I am not a doctor and nothing here is medical advice.

Written by Anna Katherine Page, who has used an LED mask three to four nights a week for two years. Last updated September 4, 2026.

How often should you use red light therapy on your face?

Two to three times a week. That is what nearly every published protocol uses, and it is the number I would give a friend without hesitation.

The reason this question is confusing is that device manufacturers recommend more. INIA suggests three to five times a week. Some brands say daily. Meanwhile the FDA clearance filed by the factory that builds INIA’s masks recommends “10 minutes a day, 3 times per week,” which is more conservative than the brand’s own marketing.

Here is what the research actually used.

Notice how narrow the real range is. Ten to twenty-one minutes, two or three times a week. Nothing in that table supports twice a day, and only one row supports five times a week.

Does three times a week work better than twice?

No, and this is the most useful single finding in the literature for anyone who owns a mask.

Bragato and colleagues, publishing in Lasers in Medical Science in 2025, randomised 95 women aged 45 to 60 into three arms: a red LED mask three times a week, the same mask twice a week, and a sham mask twice a week. Double-blind, sham-controlled, four weeks.

On the blinded specialist-rated Wrinkle Assessment Scale, there were no significant differences among any of the groups. Software-measured wrinkle length improved at the glabella and right periorbital sites. Satisfaction was high in both treated groups. The authors concluded that two weekly sessions appear sufficient.

I want to be straight about what that study shows, because it cuts both ways. The blinded human assessors could not see a difference between the treated women and the sham group at all. That is the best-designed facial trial in this field and its primary blinded outcome was null. It is also the trial that tells you three sessions buy you nothing over two.

How long should each session be?

Ten minutes for a consumer mask. That is what the devices are timed for and it is inside the published range.

The arithmetic matters more than the minutes. A mask putting out 30 mW/cm² for ten minutes delivers about 18 J/cm², which sits at the top of the 1 to 20 J/cm² band that Huang and Hamblin describe as standard for stimulation in photobiomodulation. Doubling the session to twenty minutes puts you at 36, which is outside it.

Most masks have a built-in ten minute timer and shut themselves off. That timer is not a limitation to work around. It is the dose.

Can you do red light therapy twice a day?

You can physically do it. I would not, and here is the specific reason.

Photobiomodulation follows what researchers call a biphasic dose response. Huang, Sharma, Carroll and Hamblin laid this out in Dose-Response in 2011, and their summary is blunt: “Within a decade of the introduction of LLLT in the 1970s it was realized that more does not necessarily mean better.”

The laboratory numbers behind that are striking. Fibroblasts irradiated at 632.8nm showed maximum proliferation at 2.5 J/cm² and cell damage at 10. Mouse fibroblasts at 810nm showed ATP production rising at 3 J/cm² and falling at 30. In a mouse traumatic brain injury model, the dose that worked as a single treatment, repeated daily for fourteen days, produced benefit for four days, nothing by day five, and by day fourteen the treated animals performed worse than untreated controls.

Two honest caveats. Every one of those inversion points comes from cells or rodents. There is no published human facial dose-ranging study establishing where the down-slope begins in skin, and the 2025 review says so directly: “at present, no consensus exists on optimal dosing parameters, including frequency and duration of use.”

So the case against twice a day is not that somebody proved it harmful in human faces. It is that nobody has established an upper bound, the underlying dose curve is known to invert, and the one trial that compared two frequencies found the lower one was enough. Doubling your exposure buys you no demonstrated benefit against an uncharacterised risk. That trade is bad.

Why doesn’t more red light therapy mean better results?

Because intensity is a separate variable from total energy, and this is the part that almost never gets explained.

Two experiments in the Huang and Hamblin review make the point cleanly. In a hamster oral mucositis model, 0.9 J/cm² delivered at 55 mW/cm² over 16 seconds reduced severity. The identical 0.9 J/cm² delivered at 155 mW/cm² over 6 seconds did nothing. In a rat wound model, 5 J/cm² at 4 mW/cm² improved tensile strength, and the same 5 J/cm² at 15 mW/cm² lost the benefit entirely.

Same joules. Different result. Which means “total joules” is not a sufficient description of a dose, and a brighter panel used for a shorter time is not equivalent to a dimmer one used for longer.

This is also why the LED mask marketing arms race over output figures is not the straightforward good thing it looks like. A higher irradiance number is not automatically better. It changes where you sit on a curve that nobody has mapped for human faces.

How long before you see results?

Four weeks at the absolute earliest. Six is more realistic, and that is what the FDA-cleared indication for this device class assumes.

Nothing credible in the published literature reports meaningful objective change before about four weeks. Bragato ran four. Lee ran four. Wunsch ran thirty sessions across roughly fifteen weeks. The acne literature, where blue light separates from comparators, shows that separation emerging at weeks four to eight with peak effect at twelve.

If you see a two-week before-and-after on social media, you are looking at lighting, hydration and camera angle. I say that as someone who posts before-and-afters. It is genuinely hard to shoot an honest one and almost nobody does.

There is one more gap worth knowing: durability after you stop is essentially unstudied. Cochrane’s criticism of the acne light literature was that fewer than half the trials assessed anything beyond eight weeks. Nobody can tell you what happens three months after you put the mask in a drawer.

What does a realistic routine look like?

Here is mine, after two years.

Three nights a week, ten minutes, on clean dry bare skin with nothing underneath. No serum, no oil, no SPF, no makeup, because anything on your face is between the light and your skin. Eye pads in. I do it lying down before bed, though the wireless mask I use means I have also done it while answering emails, which is the honest reason I have kept it up rather than any discipline on my part.

Moisturiser goes on a few minutes afterwards. I do not layer a retinoid on the same night, which is a caution based on barrier irritation rather than a documented interaction, and I explain the reasoning in whether you can use retinol or tretinoin with red light therapy.

The mask I use is the INIA GLOW, which measures about 30 mW/cm² independently, the same as a mask costing three times more. I wrote up what does and does not verify about it in my full two-year review.

Who should skip red light therapy for wrinkles?

Anyone who has not tried a retinoid first. Topical tretinoin for photodamage has a systematic review and meta-analysis of randomised trials behind it. LED has nothing of comparable weight for wrinkles. If your budget only stretches to one thing, the evidence points at the retinoid.

Anyone with melasma or a pigmentary disorder. Visible light is an established melasma aggravator and heat-generating devices are a recognised risk. This is a dermatologist conversation.

Anyone on photosensitising medication, or with a photosensitive dermatosis. The LED literature specifically flags screening for this.

Anyone who wants a result in two weeks. It does not happen in two weeks and a device that promises it is lying to you.

Frequently asked questions

Related reading: my two-year INIA mask review · does red light therapy work for dark spots · how to choose an LED mask on Amazon · the retinol serum I use

Routine building

Some links in this post are affiliate links. If you buy through them I earn a commission at no extra cost to you. I bought both the mask and the serum I use with my own money. I am not a doctor and nothing here is medical advice. If you are on prescription tretinoin, your dermatologist’s instruction beats anything on this page.

Written by Anna Katherine Page, who uses a retinol serum and an LED mask in the same routine and went looking for the rules. Last updated September 4, 2026.

Is there any research on using retinol with red light therapy?

No. I want to be direct about that, because the internet will tell you otherwise with great confidence.

I searched for clinical trials combining topical retinoids with LED photobiomodulation and found none. I searched for dermatology guidance and found none. What I found instead was device retailer blog content, all of it stating rules in a confident tone, none of it citing anything.

That means the sequencing advice circulating online, “retinol at night, LED in the morning,” “wait twenty minutes between them,” “never on the same day,” is not a finding. Somebody wrote it, it got copied, and now it reads like consensus. It might well be sensible. It is not evidence, and you should know which one you are getting.

Does tretinoin make you photosensitive to LED light?

This is where the confusion comes from, and the answer turns on one word: ultraviolet.

The RETIN-A MICRO prescribing information says, in the photosensitivity section: “RETIN-A MICRO can cause photosensitivity. Advise patients to avoid or minimize unnecessary exposure to ultraviolet (UV) light, including sunlight and sunlamps.” The patient section adds: “You should avoid sunlamps, tanning beds, and ultraviolet light during treatment.”

Every one of those warnings names UV.

LED face masks emit in roughly 415 to 850nm. That is visible light and near-infrared. There is no UV in it. So on a plain reading, the label’s photosensitivity warning does not apply to an LED mask.

So what is the actual risk?

Barrier irritation. It is a much more ordinary concern than photosensitivity and it is the one I would actually plan around.

Retinised skin is more reactive. The tretinoin label warns that concomitant use of other topical products “may increase the risk for local skin irritation,” and that “weather extremes, such as severe wind or cold, may increase the risk of skin irritation.” A face that reacts badly to wind is a face that may react to ten minutes of light and gentle warmth from a silicone mask pressed against it.

There is some real-world signal on the irritation side. FDA’s adverse event database for this device class contains a report from June 2026 describing a “painful, itching, burning bright red rash” on the cheeks after LED mask use, requiring high-dose antihistamines and steroids and taking two weeks to resolve, with a photosensitivity investigation noted. Those reports are unverified and causality is not established, which is exactly how they should be read. But irritation is clearly the failure mode people actually experience.

The one concrete caution in the LED literature itself is about screening. A review in the Journal of Clinical and Aesthetic Dermatology says it is “prudent to screen individuals with photosensitive dermatoses, or those taking photosensitizing medications.” That mostly means systemic drugs, doxycycline, some diuretics, St John’s wort, rather than a topical retinoid. If you are on any of those, that is a doctor conversation.

How do you actually combine them?

Here is what I do, offered as a reasonable routine rather than a rule, because there is no rule to give you.

Three sessions a week with the mask and retinoid on the nights in between fits neatly, because the published protocols for facial LED are two to three sessions a week rather than daily. I explain why more is not better in how often you should use red light therapy on your face.

The serum I use is the PRMR Youth Radiance Vitatinol serum, which I chose because it is approved by the brand for the under-eye area, and the mask is the INIA GLOW, which I have used for two years.

If you can only do one, which should it be?

The retinoid, and it is not close.

This is the comparison that nobody selling a light device wants to make, so let me make it.

The retinoid is harder to tolerate and much better supported. The mask is pleasant, easy and modest. If your budget stretches to one, the evidence points at the serum. If it stretches to both, the routine above is how I run them.

Who should not combine them?

Anyone in the first few weeks of a retinoid. The adjustment period is when your barrier is most compromised. Get stable on the retinoid first, then add the mask.

Anyone currently peeling, stinging or visibly irritated. Stop both, repair, reintroduce one at a time.

Anyone on isotretinoin. That is a dermatologist conversation, not a blog one.

Anyone on a systemic photosensitising medication. Doxycycline and its relatives, some diuretics, St John’s wort. This is the caution the LED literature actually names.

Anyone with melasma. Visible light is an established melasma aggravator regardless of what else is in your routine. I wrote that up in does red light therapy work for dark spots.

Frequently asked questions

Related reading: the retinol serum I use · my two-year INIA mask review · how often to use red light therapy on your face · does red light therapy work for dark spots