Short answer: Red light therapy for pain has moderate short-term evidence in myofascial neck pain, plantar fasciitis and lower-limb tendinopathy, weak evidence in knee osteoarthritis, early evidence in fibromyalgia, and a clear negative result in non-specific low back pain. Dose is the variable that decides which result you get.
Does red light therapy for pain work? It depends entirely on which pain
Photobiomodulation is sold as a general analgesic. The literature does not support that framing at all. Across six musculoskeletal conditions the evidence runs from moderate positive to clearly negative, and the differences are not random — they track how close the trials came to a defensible dose.
We have specified light rooms inside recovery suites, gyms and clinics as part of a business building heat and water facilities since 1987, and the question operators ask is always the same: what can we honestly tell a guest this does? Below is the condition-by-condition answer, with the meta-analyses named so you can check them.
The proposed mechanism, briefly
Red light at 630–660 nm and near-infrared at 810–850 nm are absorbed by cytochrome c oxidase in the mitochondrion, altering electron transport and downstream signalling. The proposed analgesic routes are reduced local inflammatory mediators, modest tissue repair effects, and possible direct action on peripheral nerve conduction. It is not a heat effect — a therapeutic laser at these settings barely warms tissue, which is what separates it from an infrared sauna.
Mechanism is not evidence. The rest of this article is evidence.
Six conditions, ranked by what the trials show
| Condition | Best available evidence | Verdict |
|---|---|---|
| Myofascial neck pain | 13 RCTs, 556 patients, meta-analysis | Moderate positive |
| Plantar fasciitis / lower-limb tendinopathy | 14 RCTs (817) and 18 RCTs, two meta-analyses | Moderate positive, short term |
| Fibromyalgia | 7 RCTs, systematic review | Promising, heterogeneous |
| Knee osteoarthritis | 10 RCTs (542) meta-analysis, plus a 127-patient RCT | Weak — beats placebo, adds nothing to exercise |
| Non-specific low back pain | 12 RCTs, 1,046 patients, meta-analysis | Negative |
| Delayed onset muscle soreness / recovery | Multiple small trials, no clean pooling | Mixed, unresolved |
1. Myofascial neck pain — the best musculoskeletal case
A 2022 meta-analysis in Lasers in Medical Science pooled thirteen randomised controlled trials covering 556 patients with myofascial neck pain. Low-level laser therapy significantly reduced pain intensity against placebo or active control (mean difference −1.29 on a 10-point scale), improved pressure pain threshold and improved right-side bending range of motion. Disability did not improve significantly (PMID 35962884).
The authors’ own recommendation is worth repeating because it is the pattern across this whole field: use it alongside manual and exercise therapy, not instead of them.
2. Plantar fasciitis and lower-limb tendinopathy
Two independent meta-analyses agree here, which is rare.
The first, in Musculoskeletal Science and Practice in 2022, covered fourteen trials and 817 patients with plantar fasciitis. Against placebo, laser therapy improved pain in the short term (0–6 weeks) by 2.3 points with zero heterogeneity and moderate-quality evidence, and added a 2.0-point improvement on top of conventional rehabilitation. Short-term disability did not improve, and against extracorporeal shockwave therapy there was no significant difference (PMID 34847470).
The second, in BMJ Open in 2022, is the more interesting paper because of how it was designed. It pooled eighteen trials of lower-extremity tendinopathy and plantar fasciitis and — critically — split them into dose subgroups based on the World Association for Laser Therapy treatment recommendations. Overall pain fell by 13.15 mm on a visual analogue scale at completion and remained lower four to twelve weeks later. Within the recommended dose range, the effect against placebo was larger: 14.98 mm at completion, and 18.15 mm when added to exercise therapy rather than exercise alone. No adverse events were reported (PMID 36171024).
That is the clearest published demonstration that dose, not the modality, is what is being argued about.
3. Fibromyalgia — early but consistent
A 2026 systematic review in Lasers in Medical Science identified seven randomised controlled trials and reported consistent short-term reductions in pain intensity and improvements in quality of life, with additional signals on sleep quality and psychological well-being. The authors describe it as a safe non-pharmacological adjunct and are explicit that substantial heterogeneity limits generalisability (PMID 42334638).
Seven trials is not a settled question. But for a condition where the pharmacological options are poor, a safe adjunct with a short-term signal is a legitimate thing to offer — provided it is offered in those words.
4. Knee osteoarthritis — the honest disappointment
A 2024 systematic review with meta-analysis in Physical Therapy pooled ten placebo-controlled trials and 542 participants. Pain at rest improved significantly against placebo, a moderate effect — but every included study carried unclear to high risk of bias, the Timed Up and Go test showed no effect, and the authors graded the certainty of evidence as very low and declined to recommend isolated use (PMID 38775202).
The decisive test came the year before. A six-month double-blind placebo-controlled trial in the Brazilian Journal of Physical Therapy randomised 127 knee osteoarthritis patients into exercise, exercise plus active photobiomodulation, and exercise plus placebo photobiomodulation, treating three times a week for eight weeks. All three groups improved clinically. None outperformed the others at any follow-up point (PMID 37572382).
Light may be better than nothing for a knee. It was not better than a strengthening programme, and it did not add to one. Any facility selling light for knee pain should know that trial exists.
5. Non-specific low back pain — a clean negative
A 2020 systematic review in the Journal of Physiotherapy pooled twelve randomised trials, 1,046 participants, most at low risk of bias. Against sham, the effect on pain and disability was clinically unimportant in both acute and chronic non-specific low back pain, and no better than exercise. The conclusion is unambiguous: current evidence does not support the use of photobiomodulation for this condition (PMID 32680739).
This is the most commonly claimed indication in consumer marketing and the one with the strongest evidence against it. If a supplier leads with back pain, that tells you what they have read.
6. Muscle soreness and training recovery
Unresolved. Individual trials report faster recovery of torque and reduced soreness; others find nothing; protocols vary too much to pool with confidence. If recovery is the actual objective, the evidence behind contrast therapy, sauna recovery and cold immersion is currently better developed and considerably cheaper to install.
Why dose decides everything
The reason these results scatter was described in 2009 and has not been improved on. Photobiomodulation follows a biphasic dose response: below a threshold nothing happens, within a window you get the effect, and above that window the benefit disappears again (PMC2790317, Dose-Response). More is not better. More is a different treatment.
Practically, four numbers define a session:
- Wavelength — 630–660 nm for superficial tissue, 810–850 nm for anything deeper than skin. Joint and tendon work needs near-infrared.
- Irradiance — mW/cm² at the treatment distance, not at the emitter face. Falls off sharply with distance.
- Fluence — irradiance × time, in J/cm². At 40 mW/cm² (0.04 W/cm²), a 10 J/cm² dose takes 250 seconds.
- Frequency — most positive protocols run two to three sessions a week over four to eight weeks, not daily indefinitely.
The World Association for Laser Therapy publishes recommended dose ranges by condition, and the BMJ Open analysis above showed that trials inside those ranges produced larger effects than the pooled average. A clinic that cannot state its dose is not delivering a treatment; it is delivering an exposure. That is the first thing we work through when specifying a red light therapy clinic.
Two complications worth knowing
Skin tone changes the delivered dose. A 2026 narrative review in the São Paulo Medical Journal found that melanin absorbs photons and reduces penetration in darker skin, while clinical protocols almost never adjust for pigmentation and very few trials report outcomes by skin tone (PMID 41983886). In the Gulf, where a single facility may serve every skin type in a day, a fixed protocol is quietly delivering different doses to different guests.
Blinding is genuinely difficult. Red light is visible. Sham devices are either obvious or emitting a low dose, which is not a placebo. Pain outcomes are subjective. That combination means effect sizes in this literature should be read with more caution than in a drug trial, and it is part of why “moderate positive” is the strongest verdict anywhere in the table above.
What this means if you are building or buying
For clinics and recovery facilities. Targeted, articulated applicators are the format the evidence was generated with — a therapist positioning a head over a specific tendon for a specified time. Whole-body panels are a different product answering a different demand. Both have a place; only one matches the trials.
For the treatment protocol. Write it down: wavelength, distance, time, sessions per week, total course. Then it can be audited, adjusted and defended. Almost no commercial operator does this, which is also the easiest way to differentiate.
For positioning. Light is an adjunct. Every positive meta-analysis in this article says so in its own conclusion. It works best bolted onto exercise therapy or a wider recovery circuit — light alongside cryotherapy and PEMF, or after a session in the thermal suite. In a longevity centre it earns its floor area as part of a circuit, not as a standalone claim.
For home buyers. A consumer panel used at an unmeasured distance is unlikely to replicate a clinical protocol. Our notes on red light therapy at home and the wider red light therapy guide cover what to check. And combining light with cabin heat, as in a red light sauna, compromises both functions.
Safety
The safety record is good. The BMJ Open pooling of eighteen trials reported no adverse events, and this is consistent across the pain literature. The real precautions are eye protection — near-infrared is invisible, so the pupil does not constrict against it — avoiding direct treatment over known malignancy, and caution over the thyroid and in pregnancy. Compared with pharmacological analgesia, the risk profile is a genuine advantage.
Frequently asked questions
How long does red light therapy take to work for pain?
Most positive trials measured outcomes after four to eight weeks of two to three sessions per week. Short-term effects at 0–6 weeks are where the plantar fasciitis and neck pain evidence is strongest. Single sessions are not what produced these results, and effects that appear immediately are more likely attention and warmth than photobiomodulation.
Is red light therapy better than ice or heat for pain?
Not demonstrably. In plantar fasciitis it performed no better than extracorporeal shockwave therapy, and in knee osteoarthritis and low back pain it did not outperform exercise. The consistent finding across positive meta-analyses is that it works as an addition to standard care rather than a replacement for it.
What wavelength is best for pain relief?
Near-infrared, roughly 810–850 nm, for anything deeper than skin — joints, tendons and muscle. Red at 630–660 nm is absorbed superficially and is the band behind skin and wound work. Trials in oral mucositis mostly used 660 nm because the target tissue was surface mucosa; a knee is not a surface.
Can red light therapy make pain worse?
Reported adverse events are rare and the pooled trial data report none. The more realistic failure is a null result from overdosing: because the dose response is biphasic, exceeding the effective window loses the benefit rather than increasing it, so longer sessions are not a route to better outcomes.
Specifying a room
If you are planning light therapy inside a clinic, gym or spa, the decisions that matter — treatment distance, applicator format, electrical load, cooling and where the room sits in the guest circuit — are made at layout stage, before any device is chosen. Talk to us about the room, and we will help you write a protocol you can put in front of an informed client.
