Laser acupuncture stimulates acupuncture points with a beam of low-intensity light instead of a needle. A 2004 review by Whittaker defines it as stimulation of traditional points with low-intensity, non-thermal laser irradiation: the beam does not heat the skin. A 2019 review by Tony Chon and colleagues at the Mayo Clinic notes that patients typically feel nothing during it, at most a tingling or a light touch. Because there is nothing to feel, a device can be switched off without the patient knowing, and trials have used devices that also hide from the practitioner whether the beam is on.
Wavelength, power and dose
The Chon review reports that laser acupuncture mostly uses visible red light and infrared, between 625 and 1,000 nanometres, because skin pigment and blood absorb shorter wavelengths, and that light between 650 and 900 nm passes through skin best. Lasers used on points typically put out 5 to 499 milliwatts; above 500 mW a laser is Class 4. The review adds that no guidelines exist for the parameters or the energy to deliver.
Dose is the energy delivered to each point, in joules: power multiplied by time, so a 20 mW beam held for 10 seconds delivers 0.2 J. Some trials also give energy density, joules per square centimetre of skin. A 2014 trial by Glazov and colleagues shows the scale: it used a 20 mW, 840 nm laser and gave 0.2 J or 0.8 J per point, against a sham at 0 J. Whittaker’s review notes that published studies seldom report wavelength, irradiance and beam profile in full, and that how deep the energy reaches also depends on the thickness, age and pigmentation of the skin.
The Chon review dates laser acupuncture to the 1970s, when it was used for asthma and high blood pressure in the Soviet Union. AcuiQ records how each protocol was delivered as its modality, and a protocol delivered by laser is filed as laser, separately from its points; Reading a protocol explains the field.
The blinding advantage
A needle can be felt, so a sham needle has to imitate the feeling; Sham acupuncture sets out the problems that raises. A laser at these powers gives no sensation to imitate. The Glazov trial used laser devices modified for double-blind research, and the 144 participants with chronic low back pain, the therapists and the person entering data all remained unaware of which device was emitting. A 2014 trial in JAMA by Hinman and colleagues blinded both participants and acupuncturists to laser and sham laser, and ran a needle arm and a no-treatment arm beside them, so the same trial measured laser against a sham and against nothing.
What trials and reviews report
The Hinman trial randomised 282 people aged 50 or over with chronic knee pain to needle acupuncture, laser, sham laser, or no acupuncture, for 12 weeks. Neither laser nor needles improved pain or function more than sham laser. Against no acupuncture, both improved pain modestly at 12 weeks (laser by 0.8 points on a 10-point scale), and the improvement was gone at one year. The Glazov trial found no difference in pain or disability between sham and either laser dose at six weeks; pain fell in all three groups.
A 2013 trial by Quah-Smith and colleagues in Sydney randomised 47 adults with major depression to laser or placebo laser at LV14 (Qimen), CV14 (Juque), LV08 (Ququan), HT07 (Shenmen) and KD03 (Taixi), 12 sessions over eight weeks. By intention to treat, 72.0% responded with laser and 18.2% with placebo. Clinician-rated scores improved more with laser; self-rated scores improved in both groups and did not differ.
| Review | Condition | Trials, participants | Comparison | What it found |
|---|---|---|---|---|
| Baxter 2008 | Any | 18 trials | Laser as the main treatment, against controls | Evidence supporting laser acupuncture for myofascial pain, nausea and vomiting after surgery, and chronic tension headache. |
| Law 2015 | Musculoskeletal pain | 49 trials | Control interventions | 31 of the 49 reported positive effects, and these were the trials of higher quality that reported their dose; moderate-quality evidence of benefit at an appropriate dose, seen at long-term follow-up rather than straight after treatment. |
| Glazov 2016 | Chronic low back pain | 15 trials, 1,039 | Sham laser, with blinded assessment | Pain up to 1.4 cm lower on a 10 cm scale in the short term, but only in trials giving at least 3 J per point, in people with pain for under 30 months, and in trials that did not use acupuncture points; moderate-quality evidence. |
| Zeng 2025 | Knee osteoarthritis | 5 trials, 293 | Placebo laser; exercise | Pain 2.33 cm lower on a 10 cm scale than with placebo laser; effect against exercise uncertain; low-certainty evidence. |
| Kamau 2026 | Knee osteoarthritis | 13 trials, 611 | Sham, placebo, electroacupuncture or standard care | No significant improvement in pain, function or knee bending; wavelengths ranged from 650 to 904 nm and heterogeneity was high. |
| White 2014, Cochrane | Stopping smoking | 2 trials of laser | Various | Results inconsistent and could not be combined. |
AcuiQ’s pages for knee pain, chronic low back pain, depression and myofascial pain list the protocols individual studies prescribed, each with its citation. Knee osteoarthritis and Low back pain report the needling evidence for the same conditions.
Safety
The Chon review classes acupuncture lasers as Class 3b, which can damage the eye, and states that patients and practitioners need protective eyewear. It lists precautions over a fetus, over the heart in people with heart disease, over areas of bleeding and over the gonads, and against treating children under 2, and it reports transient dizziness, headache and fatigue as adverse effects. In the Quah-Smith trial transient fatigue was the only adverse effect reported, and the Hinman trial reported no serious adverse events. The needling safety review left laser out, so its rates do not apply here.
What the sources cannot tell you
A negative laser trial may have tested too little energy. The Glazov meta-analysis found benefit only at 3 J per point or more, while the Glazov trial gave at most 0.8 J; Law and colleagues found that the negative or inconclusive trials in their review commonly were of lower quality or did not report their dose adequately. Because trials differ in wavelength, power, time and points, and many do not report them fully, two trials of “laser acupuncture” may not have delivered comparable treatments. The low-back meta-analysis also found its effect in laser trials that did not use acupuncture points, so it does not show that the effect depends on stimulating a point.