The randomised evidence says low-level laser therapy does not work for chronic tinnitus: a 2022 review found ten randomised trials, and eight reported it ineffective. Reviews that counted weaker studies reached warmer conclusions.
Low-level laser therapy — usually called photobiomodulation in the research literature — is one of the more heavily marketed tinnitus treatments. Four systematic reviews have looked at it and reached four different conclusions.
That disagreement is the most useful thing on this page, because the reason for it is visible.
Start with the randomised trials only
A 2022 systematic review in European Annals of Otorhinolaryngology, Head and Neck Diseases asked a deliberately narrow question: what do the randomised controlled trials show? It found ten of them — two single-blind, eight double-blind.
Eight of the ten reported that photobiomodulation was ineffective for chronic tinnitus. One reported benefit lasting up to three months. One reported resolution figures of 30% with a diode laser and 100% with a Neodymium-doped Yttrium Aluminium Garnet laser — a result so far outside what the rest of the field produces that it belongs in the patterns that make a weak result look strong rather than in a summary of what laser therapy achieves.
The review also graded the trials: risk of bias was high in four, medium in five, and low in one. Its conclusion is that the effectiveness of photobiomodulation remains debatable and that properly powered trials with long-term follow-up are still needed.
Then widen the net, and the answer changes
A 2023 systematic review in The Journal of Laryngology and Otology searched more broadly and found 28 studies reporting outcomes in 1,483 people, plus two animal studies. Its conclusion: outcomes are "generally positive and superior to no photobiomodulation therapy", while evidence of long-term benefit is deficient.
Two reviews of the same treatment, two opposite headlines. The difference is what each counted — one admitted only randomised trials, the other admitted the wider literature. It is a recurring pattern: uncontrolled studies show a bigger effect than controlled ones, reliably enough that the gap between them is a measurement of bias rather than of treatment.
A 2021 review in the Journal of Speech, Language, and Hearing Research sits in between. It found seven randomised trials, judged one at low risk of bias and six at moderate, and reported that all seven found significant results of some degree — but with no consensus among them, and it declined to make any recommendation about clinical use.
The most recent, a 2025 review in Cureus covering nine clinical studies, reports short-term reductions in tinnitus severity with several studies beating placebo, and adds the durability caveat directly: those benefits often diminished after three to six months.
"Low-level laser therapy" is not one treatment
The 2023 review recorded the parameters the 28 studies used. They span ten different wavelengths, with sessions running from nine seconds to thirty minutes, and follow-up from seven days to six months.
That range is worth pausing on. A drug trial comparing doses that differed by a factor of two hundred would not be pooled into one verdict, and nobody would expect a single answer from it. When reviews of laser therapy disagree, part of the explanation is that they are not reviewing the same intervention — and part of why no clinic's protocol can claim the support of "the research on laser therapy" is that the research has not converged on a protocol.
The trial this page used to rest on
For a long time this article described a single study as the strongest available evidence, and it is worth keeping because it shows how a positive trial can still be a discouraging one.
A double-blind randomised trial examined laser therapy in patients whose tinnitus accompanied noise-induced hearing loss — all male workers aged 30 to 51, average tinnitus duration about two years — across 20 sessions delivered every other day, 20 minutes each. Against placebo it improved all three outcomes: a visual analogue scale, the Tinnitus Handicap Inventory, and tinnitus loudness.
Then the details. The non-response rate was 57% immediately after treatment and 70% at three months, and the authors describe the effect as fading after three months, calling the pattern "U-shaped" over follow-up.
So the positive trial reports that most participants did not meaningfully improve, and that those who did largely stopped having improved. Set beside eight randomised trials finding nothing, it reads less like the strongest evidence for laser therapy than like the high-water mark of it.
A newer device, a smaller trial
A 2025 randomised, single-blind, placebo-controlled trial in Clinical Otolaryngology tested a specific device, the TINI laser, in 38 participants across three arms: eight active treatments, four active followed by four sham, or sham only.
It found significant improvement in the emotional subscale of the Tinnitus Handicap Inventory and in loudness and perception scores in the fully active arm. The partly active arm improved during active treatment and those effects diminished once the sham sessions began. Its authors conclude that intermittent photobiomodulation may be effective and that long-term efficacy and safety still need exploring.
Thirty-eight participants, single-blind, one device. It is a reason to run a larger trial, not a reason to buy a course of treatment.
What to do with this
It is not dangerous. Every review reports laser therapy as safe and well tolerated, and none raises a safety signal. The argument against it is about money and expectation, not risk.
The randomised evidence does not support it. Eight of ten randomised trials found nothing; the reviews that read more warmly did so by counting weaker designs, and the one that stayed strict declined to recommend it.
If it is offered to you, the useful questions are what wavelength and session length the clinic uses, what evidence exists for that protocol specifically, and what the total cost of a course is. Given the parameter range in the literature, a clinic pointing at "studies on laser therapy" is usually pointing at studies of something other than what it is selling.
Where the guideline lands
The 2024 VA/DoD tinnitus guideline suggests against low-level laser therapy for tinnitus management (weak against). It is one of the few treatments that guideline comes out against rather than calling the evidence insufficient, and it reaches the same place as the randomised trials above.
How it compares
Compared with what the evidence can say about CBT or hearing aids where hearing loss is present, laser therapy is not a close call. How this site rates evidence sets out the standard being applied.
A note on what this page used to say
An earlier version of this article rested on that single 2013 trial and concluded it showed "a real effect rather than a placebo response". That was a fair reading of one trial and a poor description of the literature, which contains ten randomised trials, eight of them negative. The corrections log records the change.