How Often Should You Do Red Light Therapy? What the Trials Actually Used

Two to five sessions a week covers almost every published red light protocol, with a handful of daily ones, and there is no evidence that the top of that range does more than the bottom. The honest reason to pick a frequency is not that it is optimal. It is that trials run for months, not days, and the only variable you fully control is whether you are still doing it in week twelve.

That sounds like a cop out. It is actually the most defensible thing anyone can say about frequency, because of a fact the category never mentions: exactly one human trial has ever compared two treatment frequencies head to head, and it found no difference between them.

The only frequency comparison that exists

A 2025 randomized, sham controlled, double blind trial in 95 women aged 45 to 60 was built specifically to test application frequency in facial rejuvenation. One arm used a 660nm mask twice a week, another used it three times a week, both for four weeks. That is 8 sessions against 12, a 50 percent difference in total exposure.

Nothing separated them. Participant satisfaction came in at 73.4 percent for the two per week arm and 79.6 percent for the three per week arm, with no significant difference between the two active groups, and the photographic assessments taken 30 days after treatment did not separate them either.

Two things have to travel with that result, and marketing pages carry neither.

First, the same trial found no significant difference from sham on the clinician rated Wrinkle Assessment Scale at all. Software based measurement did find reductions in wrinkle length, and patient satisfaction was higher in both light arms than in sham, so it is mixed rather than flatly negative. But a frequency comparison nested inside a trial whose primary outcome was null is a weak place to build a schedule from.

Second, it tested 2 against 3 over four weeks. That is the entire human evidence base on frequency. It says nothing about 1 versus 5, nothing about daily, and nothing about what happens past a month.

So when a device manual specifies a frequency, or a comparison page tells you the optimal cadence, understand what that number is: a convention, or a manufacturer instruction, copied forward. Trial schedules in this literature were chosen, not compared.

Is daily better than three times a week

Nobody knows. There is no human randomized trial comparing daily use to alternate day or three times weekly dosing, for skin, hair, recovery, pain or sleep. Anyone who tells you daily is better is telling you what they think, not what was measured.

What does exist points weakly in the other direction, and it is worth knowing how weak it is. In a mouse traumatic brain injury model, animals given 14 consecutive daily laser treatments appeared to do worse from day 16 to day 28 than animals that got no treatment at all. The authors state plainly that those differences were not statistically significant. That is a hypothesis in mice, not a finding of harm in people, and it should never be presented as one.

One piece of provenance travels with that observation and with everything in the next paragraph. All of it, the mouse frequency trend included, comes from a single 2011 narrative review of animal and cell experiments written by four of the field's most prolific proponents. These are not independent lines of evidence converging on an answer. They are one paper, and none of it is primary human data.

That mouse observation sits alongside the biphasic dose response, the pattern most often invoked in this field: too little light does nothing, an intermediate dose produces an effect, and a higher dose progressively loses the benefit. That shape was characterised in cell culture and in animals, and the review authors themselves note there has been no convincing report of it occurring in patients. It is repeated far more consistently than it has been demonstrated: the one human trial that put three different dose levels against sham found no difference between any of them, which is not what a steep dose response predicts. It is also reached for as a way of explaining away trials that found nothing, and it does not license that. Where the top of the curve sits in human tissue is unknown for skin, muscle and scalp alike, so more frequent exposure cannot be assumed to be more.

Put those together and the defensible position is narrow: more frequent is not automatically better, more frequent has never been shown to be worse in humans, and treating daily use as an upgrade is an assumption. If you want the reasoning behind why a longer session is not a free upgrade either, that is the session length page, and this one will not re-derive it.

BEYOND SUPPLEMENTS No stack adds muscle you did not earn. EMS might. Twenty minutes, twice a week, every major muscle group under load. Here is what the studies actually measured. READ THE RESEARCH →

Weekly structures by goal

These are not recommendations from a guideline. The only guideline grade dosing protocols that exist anywhere are clinician delivered intra-oral laser for preventing oral mucositis, in stem cell transplant patients through chemotherapy conditioning and in head and neck cancer patients through a radiotherapy course. They are aimed at spots of 0.04 to 1 cm2 at irradiances up to 1000 mW/cm2, roughly 20 to 200 times what a consumer panel puts out, and they do not transfer to a panel or a mask.

One other guideline touches this field. The European evidence based S3 guideline for pattern hair loss suggests low level laser therapy only as ancillary therapy, using devices at energy levels shown to work in randomized trials, and states that it cannot make a recommendation for or against treatment beyond six months. Note what it does not do: it names no frequency, no session length and no course length. Nothing covers skin, athletic performance or general wellness at all. What follows is what the trials actually ran.

Skin. The sham controlled trial of a home use LED mask used 630nm red plus 850nm infrared for 9 minutes, five times a week for 12 weeks, which is 60 sessions. The treated and sham groups did not separate at week 4. Significant differences in crow's feet grading appeared at weeks 8, 12 and 16. Three things travel with it that a product page will not tell you: it was funded by the mask's manufacturer, its sham device still emitted 630nm at one tenth intensity rather than no light at all, and 59 people completed it in a single country. The frequency trial above used 21 minute sessions twice or three times weekly for 4 weeks and found no difference between those two cadences.

Against both of those sits a split face trial in 137 women that reported a 31.6 percent periocular wrinkle volume reduction after 10 sessions over 4 weeks. That design put active light on both sides of every face, so its figure is a change from baseline rather than a difference against a control, and it drew a published methodological rebuttal in the same journal. It must never be quoted alone as a four week result.

Hair. A systematic review of seven double blind sham controlled trials of home devices in 607 participants recorded the schedules used: about 3 to 4 times a week for most devices and daily for one, with sessions from 90 seconds to 36 minutes. A second review of ten trials, pooling substantially the same small set of studies rather than confirming them independently, records sessions of 15 to 30 minutes delivered anywhere from daily to three times a week. The two do not agree on session length, and that disagreement is itself the finding: session length was never a controlled variable. Every one of those trials ran 16 to 26 weeks. This is the goal where the frequency question matters least and the course length question matters most, because the measurement endpoints are half a year out.

Pain. Across 14 randomized trials, courses ran from a single session to 20 sessions delivered over anywhere from 3 days to 12 weeks, with 2 to 3 sessions per week the most common pattern. Follow up assessment usually ended within 8 weeks of the final session, so almost nobody has measured how long any benefit lasts after you stop. The one exception is a manufacturer funded fibromyalgia trial in a light bed that reported improvement six months after the final session on outcomes that were all self reported questionnaires. Note also that nearly all of this was a clinician pressing a contact laser probe onto specific points, which is a different thing from standing in front of a panel.

Recovery. Frequency is the wrong frame here, because the sessions attach to training rather than to a calendar. The evidence here splits along delivery format rather than along schedule, and both halves have to be stated. Positive results exist: pooled across trials, light before resistance exercise added about four repetitions, improved strength recovery by a standardized mean difference of 0.24, and reduced soreness after a laboratory muscle damage protocol. Every one of those came from a laser or LED cluster held in contact with a specific muscle, which is not the format most people own. On the other side, the only systematic review of whole body panels found that none of its five studies reported any benefit to performance or fatigue, though two of the five reported better sleep quality and the review pooled nothing statistically, so that is an absence of demonstrated benefit rather than proof of no effect.

The trials that ran for weeks rather than a single session found nothing extra. A blinded trial of 60 joules, 300 joules and an escalating dose across ten strength sessions found no differences from sham on peak torque, total work, perceived exertion, pain or blood lactate, and six weeks of light added to sprint and squat training in trained men produced no benefit over placebo. Both were small, at twelve to thirteen people per group in the first and 39 across three arms in the second, so neither excludes a small effect. For where in the day to place a session around training, see the timing page.

Why frequency is the wrong number to optimize

Look down the table and the pattern is obvious. Frequency varies between 2 and 7 sessions a week across the entire literature and nobody has shown it changes anything. Course length is set by the tissue, not by the schedule, and where each tissue runs out of evidence is set out in the results timeline.

This is the practical reason frequency beats duration as a thing to optimize. A schedule you sustain for 16 weeks at three sessions a week delivers 48 sessions. A schedule you sustain for 3 weeks at seven sessions a week delivers 21 and then stops. The first one is inside the range where trials measured something. The second is outside every skin and hair protocol that has been published.

Two honest limits on that reasoning. Nobody has run a trial testing adherence, so "consistency matters" is an argument from what the trials happened to do, not a demonstrated dose response. And nothing in the literature tells you what a week should look like once a course is finished, because no maintenance schedule has ever been tested.

Can you make up a missed session

Not reliably, and the reason is more interesting than the answer. Total joules do not define a dose. In a rat wound study, an identical total dose of 670nm light produced a significant improvement when it was delivered slowly at a low intensity and lost the effect entirely when the same energy was delivered faster at a higher intensity. In a hamster model, the same energy per point reduced mucositis severity at one intensity and did nothing at a higher one.

Both of those are animal experiments, and both are reported inside the same 2011 narrative review cited earlier, so they are two results summarized by one advocate authored paper rather than two independent findings waiting to be confirmed. Neither has a human counterpart. This is not a rule. It is a caution against the obvious shortcut: missing Tuesday and doubling up on Wednesday is not arithmetic, and standing closer to cut the session short is the exact substitution those two experiments failed at.

The workable version is dull. Miss a session, do the next one on schedule, and judge the block on how many sessions you completed over three months rather than on any single week.

Building one you will keep

Pick the smallest frequency inside the published range, not the largest. Three a week sits inside every weekly band in the table above and inside the only frequency trial ever run. Recovery is the exception, because those sessions attach to training rather than to a calendar.

Anchor it to something that already happens rather than to a time. Be clear that this is an assumption and not a finding: no trial in this literature reported its own adherence, so nothing is known either way about what people actually complete. A fixed slot after an existing daily habit is a reasonable bet, not a measured one.

Set the review date before you start, not during. Skin trials measured anywhere from 4 to about 16 weeks, with the one sham controlled home trial not separating from sham until week 8. Pain courses ran from 3 days to 12 weeks. Hair trials measured at 16 to 26 weeks. Deciding in week three whether a skin or hair course is working is deciding before those trials would have looked. For where each tissue runs out of evidence, see the results timeline.

And treat the schedule as one of four variables, not as the protocol. Wavelength, dose, frequency and target tissue interact, and the protocol guide walks through deriving all four. If you have not yet decided whether the evidence justifies buying anything, the evidence guide ranks what holds up by strength of evidence. If you have decided and want the hardware comparison, that is the device roundup.

Safety and who should not do this on any schedule

Frequency changes cumulative exposure, and cumulative exposure is exactly where the safety evidence runs out. Follow up in these trials generally ends within about eight weeks of the last session, no controlled trial has safety follow up longer than 26 weeks, and no guideline provides a validated contraindication list, a limit on cumulative exposure, or dose limits for repeated whole body sessions.

Within those short windows, reported side effects are mild and local: dry skin, itching, scalp tenderness, a warm sensation at the site, and mild transient erythema in four patients across seven skin rejuvenation trials, with no serious adverse events reported. That is short term tolerability, not a long term all clear, and devices sold under the general wellness policy are not required to report adverse events at all, so the silence in the reporting databases carries no information.

• Photosensitising medication. Most photosensitising drugs have an action spectrum dominated by ultraviolet A at 315 to 400nm, outside the band a red panel emits. But the same source records a subset that react to visible light from 400 to 740nm, which is inside the band a red panel emits, and almost none of these drugs have been tested against red or near infrared light directly. Individual action spectra vary, so this is not a reassurance. Two are unambiguous: porfimer sodium is activated at 630nm, inside the band most red panels emit, and its label requires avoiding sunlight and bright indoor light for at least 30 days. Verteporfin is activated at 689nm and its label instructs patients to avoid bright light for five days after infusion, naming light emitting medical devices specifically. If you take a photosensitising drug, ask the prescriber rather than a device manual.
• Light sensitive conditions. Visible light reactivity is documented in solar urticaria, cutaneous lupus and the porphyrias, but the studies did not isolate red wavelengths, so nobody can tell you these conditions are safe with a red panel.
• Pregnancy. No study of consumer LED panels or whole body red light exposure in pregnancy exists. Manufacturers list it as a contraindication by default, which is an absence of evidence rather than a finding.
• Eye protection. Use it, and use it more carefully as frequency rises. A 37 year old woman developed photochemical retinopathy after using an LED face mask emitting 460 to 470nm blue light for 20 minutes every other day for about a month with her eyes open and no protection. That case implicates the blue channel, not red or near infrared, but it is the reason a multi wavelength mask with a blue setting deserves eye protection. Separately, near infrared output gives no visual stimulus, so the eye's aversion response and pupil constriction cannot be assumed, which is why international exposure guidance sets a distinct limit for it. Whether eye protection is necessary for visible red at consumer intensities has never been established either way.
• Children, adolescents, and anyone with a scalp or skin lesion under investigation. No trial evidence exists in the first group. In the second, photobiomodulation is proliferative and one review flags dysplastic or malignant lesions as a theoretical caution.

Regulatory status does not fill any of these gaps. A 510(k) is a finding of substantial equivalence to an already marketed device, which is a clearance, not an endorsement that the device works, and a search of the 510(k) database returns no clearance on file under the brand names of the best known consumer panels, which ship instead under the general wellness policy, which the guidance itself says does not establish that a product has been shown to be safe or effective.

General information, not medical advice. Talk to a doctor before starting if you are pregnant, take photosensitising medication, or have a light sensitive condition.

FAQ

How often should you do red light therapy?

Published protocols run from 2 to 5 sessions a week, with a handful of daily ones, and for hair specifically 3 to 4 times a week for most devices. There is no established optimum: the only human trial comparing two frequencies tested 2 against 3 sessions a week over 4 weeks and found no difference between them. Pick the lowest frequency inside that range that you can hold for the length of the course.

Is daily red light therapy better than three times a week?

No human randomized trial has ever compared daily use to alternate day or three times weekly dosing, so the answer is not known. The only relevant data is a mouse study where 14 consecutive daily treatments produced a non significant trend toward worse outcomes than no treatment, which is a reason not to assume daily is an upgrade rather than evidence of harm.

How long should a course of red light therapy run before you judge it?

Judge a block on sessions completed rather than weeks elapsed. Three a week held to for four months puts you inside the range the trials used; a burst of daily sessions that stops in week three does not. The point at which each tissue stops having evidence behind it is set out in the results timeline.

What happens if you miss sessions?

Nothing is known, because no trial has tested adherence. Do not try to compensate by doubling up or by standing closer to shorten a session: in animal studies, the same total energy delivered faster at a higher intensity lost the effect that the slower, weaker delivery produced.

Do you need to keep going forever?

No maintenance schedule has been established for any indication, so there is no evidence based answer to what a post course week should look like. What is known about whether gains persist after stopping is covered in the results timeline.

Does more frequent use produce faster results?

There is no evidence for it. Photobiomodulation follows a biphasic dose response in cells and animals where benefit rises to a peak and then falls away, and the one human frequency comparison found 12 sessions no better than 8 over the same four weeks.

Is there an official recommended schedule?

Not for any consumer use. The only guideline grade dosing protocols anywhere are clinician delivered intra-oral laser for preventing oral mucositis in cancer patients, and the panel is explicit that its specific parameters should be followed as whole recipes rather than treated as interchangeable numbers. One guideline does touch a consumer indication: the European evidence based S3 guideline for pattern hair loss suggests light only as ancillary therapy and cannot make a recommendation for or against use beyond six months, but it names no frequency. Nothing covers skin, athletic performance or general wellness.