Red Light Therapy: What It Treats, Which Wavelength to Use, and How to Run a Session Safely

Yes for specific goals, no for most of the rest. Red light therapy has reasonable support for skin aging and narrow cleared use in pain relief, and little or none for eczema, fat loss, or mood. Clinically effective doses in reviewed trials ran roughly 3.8 to 126 J/cm², so the dose decides everything.

TL;DR: Key takeaways

• FDA 510(k) clearance (21 CFR 890.5500, product code NHN) covers narrow adjunctive non-thermal pain relief, not anti-aging, fat loss, or mood claims, and no such device is FDA approved for general use.
• Red light at about 630 to 700nm acts at the skin surface; near-infrared at about 800 to 850nm reaches muscle and joint, so choose by target depth.
• Effective doses in reviewed trials ran roughly 3.8 to 126 J/cm², which makes distance-and-minutes advice meaningless without your device's published mW/cm² output.
• Expect weeks, not days: skin texture at 4 to 8 weeks, pain within days to 2 weeks, hair at 12 to 24 weeks, acne still improving 12 weeks after a course ends.
• Never aim any device at unprotected eyes and never at a child's eyes; US eye-directed devices are unregulated and ophthalmologists warn they could damage a child's retina.

What red light therapy actually does inside a cell

Red light therapy, also called photobiomodulation, uses red and near-infrared LED or laser light to deliver energy into skin and deeper tissue. It contains no ultraviolet wavelengths. It is used mainly for skin and pain indications, and its effect is strictly dose dependent: the same device helps, does nothing, or irritates, depending on how much energy actually reaches the tissue.

The mechanism is reasonably well mapped. Mitochondrial cytochrome c oxidase absorbs red light in the 620 to 700nm range and near-infrared light from 700 to 1440nm, which raises ATP output and sets off downstream changes in inflammation signalling, collagen synthesis, and tissue repair. On packaging you will see the same thing sold as photobiomodulation, low-level laser therapy (LLLT), or simply PBM. The American Academy of Dermatology confirms these devices use no UV wavelengths, which is why this is not a tanning exposure and not a DNA-damaging one.

One caution before the numbers arrive. "Cells absorb light" is not the same claim as "any amount of light does something useful," because the biological response tracks the energy delivered, which the protocol section below turns into arithmetic.

An FDA clearance is not proof the device does what the ad says

Clearance and approval are different regulatory animals, and the gap matters when you are about to spend a few hundred dollars. Through the 510(k) pathway the FDA clears powered light-based instruments for non-thermal pain therapy (21 CFR 890.5500, product code NHN), which means a device was judged safe and substantially equivalent to something already on the market. Those clearances cover narrow adjunctive indications such as temporary relief of pain. They do not mean anyone at the agency reviewed evidence that the panel smooths wrinkles, reduces body fat, or lifts mood. Those claims simply ride along on the same box.

The second half of the reframe is about time. Benefit, where it exists, accrues over weeks of consistent and correctly dosed sessions, and it can keep building after a course finishes. It does not come from one long or high-intensity blast, and treating the device like a tanning bed is how people get irritated skin and no result.

So read "FDA cleared" on a listing as a safety-and-paperwork signal, then go do two things: find the published irradiance in mW/cm² at a named distance, and read the actual cleared indication. The broader safety picture is genuinely reassuring, with one real exception, and that exception is anything aimed at the eyes.

The benefits sort into three confidence tiers, not one list

The honest way to present red light therapy benefits is graded, because the evidence behind each marketed use is wildly uneven.

Tier 1, reasonable support: skin aging. A controlled trial published in Photomedicine and Laser Surgery found improvement in wrinkles and in collagen, which makes this the best-supported cosmetic use and the one most worth your four to eight weeks.

Tier 2, mixed: acne. A 2024 Cochrane review found no high-certainty evidence for light treatment of acne. The AAD is equally specific about what it can and cannot touch: visible light treats inflammatory pimples and does nothing for blackheads, whiteheads, cysts, or nodules, and a single session rarely produces meaningful results. It may help some people with inflammatory lesions. It is not a substitute for proven acne treatment.

Tier 2, mixed: pain and muscle recovery. Trials are positive but preliminary, and this is the one area with any FDA clearance behind it, for adjunctive non-thermal pain relief. In-clinic sessions are usually delivered inside a course of physical therapy rather than sold alone, so if cost is your real question, start with whether Medicare covers physical therapy.

Tier 3, unsupported. Eczema has no supporting evidence. Devices marketed to slow nearsightedness in children are not FDA approved and are unregulated in the US, and the American Academy of Ophthalmology points parents toward evidence-based options instead: less screen time, more outdoor time, low-dose atropine eye drops, or specialized contact lenses.

Hair loss sits awkwardly between tiers. The mechanism is plausible, the timeline is long, and the only fair way to judge it is described below.

660nm and 850nm are not interchangeable: choose by depth

The most useful thing to know before buying is that red light therapy wavelength determines how deep the energy goes, not how strong it is.

Visible red, roughly 630 to 700nm and usually sold as 630nm or 660nm, is absorbed in the epidermis and upper dermis. That is the band for collagen, texture, tone, and inflammatory acne lesions. Near-infrared, roughly 800 to 850nm and usually sold as 810nm or 850nm, carries further into muscle, tendon, and joint, which makes it the band for pain and recovery goals. Near-infrared is invisible, so a panel that looks dim may be doing more than one that looks dazzling. Apparent brightness tells you nothing about output.

Both bands sit inside the red and near-infrared absorption windows for cytochrome c oxidase, which is why the mechanism is shared and the targets are not. The practical rule: skin goal means a red-dominant device, joint or muscle goal means a near-infrared-dominant one. Dual-wavelength panels are fine, but only if output is specified per band rather than as one combined number. And infrared heat lamps and infrared saunas are a different exposure altogether, broad spectrum and thermal, not what these trials tested.

How to use red light therapy at home: the dose you can actually calculate

Dose is the step almost every home guide skips, and it is the only one that separates a session from a placebo with a power cord.

1. Find your irradiance. Get the manufacturer's mW/cm² output at a stated distance. If no figure is published at a named distance, you cannot dose, and you should treat the device as unverified.
2. Do the arithmetic. J/cm² = (mW/cm² x minutes x 60) / 1000. A panel delivering 30 mW/cm² for 10 minutes gives about 18 J/cm².
3. Aim inside the researched band. Reviewed clinical trials clustered roughly between 3.8 and 126 J/cm², lower end for skin goals, higher end for deeper pain targets.
4. Fix your distance. Use 3 to 6 inches for small handhelds and spot work, 6 to 12 inches for most panels, and 12 to 18 inches for whole-body sessions with longer run times. Irradiance drops steeply as you back away, so choose one distance and keep it.
5. Run the session. Five to 15 minutes per area on clean, dry, bare skin, the device perpendicular to the target, eye protection on, three to five sessions a week.
6. Change one variable at a time. Hold a protocol for four weeks before adjusting anything, then change time or frequency, never both, and redo the dose math afterwards.
7. Log it. Distance, minutes, and area treated, every session, so the dose is reproducible.

Here is what most pages will not tell you: plenty of consumer devices publish a wattage figure, a diode count, and nothing you can convert into a dose, and a few publish irradiance measured at the glass, where nobody treats. If the spec sheet will not give you mW/cm² at a named distance, you are guessing, and so is anyone telling you to stand 8 inches away for 10 minutes.

The habit beats the hardware. Same time, same spot, same distance is the whole discipline, because adherence is the variable home users actually fail on, not wattage. If keeping the record is the part that slips, a health companion app like August AI can help you keep that session log and your questions about it in one place.

Does red light therapy work? Only on the research's timeline, not yours

Each goal has its own window, and quitting before the window opens is the most common reason people conclude it failed.

For skin texture and wrinkles, judge at roughly four to eight weeks of consistent use, twice weekly or more, which matches the design of the controlled wrinkle and collagen trial. Pain and stiffness tend to respond faster, often within days to about two weeks, and that is also the only indication with clearance behind it. Hair regrowth needs 12 to 24 weeks before any fair assessment, because that is simply how long a hair cycle takes to show you anything.

Acne is the instructive case. In a study cited by the AAD, lesion counts fell about 50% after a four-week course, then kept improving to a 72% reduction measured 12 weeks after the final treatment. Stopping at week four and calling it a failure misreads the curve.

The stop rule matters as much as the start. If nothing has changed by your goal's timeline, re-check three things before buying another device: the published output, the distance you actually used, and the number of sessions you actually completed. If the dose was in range and the sessions were logged, the honest conclusion is that it is not working for that goal. Photograph the same area in the same light once a week rather than judging in the mirror, and a health companion app like August AI can help you track those changes alongside the session log.

Side effects are rare, but not equally rare for everyone

Serious red light therapy side effects are uncommon, and the baseline picture is reassuring: the AAD notes these devices use no UV wavelengths and that research has not found they cause any type of cancer, and considers them safe for short-term use. If you have been treated for cancer, the question deserves a direct answer: a systematic review of in vitro, preclinical, and clinical data found no evidence linking photobiomodulation to new or recurrent malignancy and no reason for people previously treated for cancer to avoid it. Clear it with your oncology team anyway.

What people do report is mild and short-lived: transient redness, warmth, a tight feeling in the skin, or a headache and eye strain from looking at the panel. Two groups need more care. The AAD warns that people with darker skin tones are more sensitive to visible light devices and face a higher risk of hyperpigmentation that can be more intense and longer lasting, and that people with light-sensitive conditions such as lupus should avoid this treatment entirely.

Check your medicine cabinet before your first session. Photosensitizing products raise your risk of a reaction:

• Isotretinoin
• Doxycycline and other tetracycline antibiotics
• Some diuretics
• St John's wort
• Topical retinoids on the area you plan to treat

Eye safety is where generic advice fails. Wear opaque goggles rather than relying on closed eyelids, never point a panel at the face without covering the eyes, and never use an eye-directed device on a child. The American Academy of Ophthalmology quotes a pediatric ophthalmologist saying there is not enough information to assess the safety of these devices and that they could damage your child's retina. That is the strongest warning on this page, and it is not softenable.

If a skin reaction appears and you cannot tell ordinary irritation from a burn, August's $39 online urgent care visit lets a US-licensed clinician evaluate it and prescribe if that is appropriate. The clinician decides. A video visit also cannot biopsy a lesion or examine your retina, so a suspicious spot or any change in vision needs an in-person exam.

Some reactions mean stop immediately, not push through

Call 911 or go to an emergency room now if:

• You have sudden vision changes, eye pain, or persistent spots or afterimages after light exposure near the eyes. This needs urgent ophthalmologic care.
• There is blistering, skin breakdown, or severe pain at the treatment site.

Book a visit within the next week if:

• You have a new or changing mole, a non-healing sore, or any suspicious lesion. It needs dermatologist evaluation before any light therapy, and must never be treated with light.
• Lupus or another photosensitive condition flares after a session.
• Redness or new darkening has not settled after 48 hours.

Talk to a clinician before you start if:

• You take a photosensitizing medication.
• You have lupus or another photosensitive condition.
• You are pregnant.
• You have a history of, or active treatment for, skin cancer.
• There is an undiagnosed lesion in the area you want to treat.
• You want to use any device on or near a child.

Unsure which tier you are in? The free August AI Symptom Checker can help you sort same-day care from a scheduled visit. It is triage, not a diagnosis.

Bottom line: dose it right, then give it eight weeks

Three moves carry this whole decision. Match the wavelength to the depth you need to reach, calculate your dose in J/cm² from your device's published output at a stated distance, then run that protocol unchanged for your goal's full timeline. Clearance is paperwork, consistency is the treatment, and the eyes are the one place to be strict. Red light therapy rewards arithmetic and patience, and if the timeline passes with nothing to show, stop and put the money toward a treatment that has been evaluated for what you actually want fixed.

Disclaimer: This article is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare provider for diagnosis and treatment decisions. If you are experiencing a medical emergency, call 911 or go to the nearest emergency room immediately.