Does TMS Work for Tinnitus? The Short Answer
Repetitive TMS (rTMS) consistently reduces tinnitus-related distress more than sham treatment in the short term, but its effect on tinnitus loudness is weak, benefits beyond six months are not well established, and no major clinical guideline currently recommends it for routine use. Two large meta-analyses (He et al. (2025); Liang 2020) confirm small-to-moderate short-term effect sizes on distress scores. A third meta-analysis found no benefit at any time point. The German S3 guideline formally recommends against routine rTMS for tinnitus, though a dissenting expert group considers it an option when other treatments have failed.
Why Patients Are Searching TMS as a Tinnitus Treatment
If you are researching TMS for tinnitus, you have probably already tried, or seriously considered, sound therapy, cognitive behavioural therapy (CBT), or tinnitus retraining therapy (TRT). Those approaches help many people. But if you are still searching, you may be looking for something that targets the neurological source of the sound rather than just helping you manage it. TMS, or transcranial magnetic stimulation, is often described as a “brain stimulation” treatment, and commercial clinic websites sometimes cite response rates of 35–50%. That framing is understandable, but it leaves out a lot.
This article is an independent evidence review. We are not selling TMS, and we are not dismissing it either. The goal is to give you what the clinic websites and the academic reviews typically don’t: an honest picture of what the research actually shows, what remains uncertain, and what practical steps make sense if you are weighing this option.
What TMS Is and How It’s Supposed to Work for Tinnitus
Transcranial magnetic stimulation uses a coil placed near the scalp to deliver focused magnetic pulses. Those pulses briefly alter the activity of neurons in the targeted area of the brain. The “repetitive” in rTMS refers to delivering pulses in sequences rather than single shots, which produces more lasting changes in how readily neurons in the targeted region fire.
For tinnitus, researchers have focused on two brain targets, each addressing a different part of the problem.
The first is the left auditory or temporoparietal cortex. The leading theory of tinnitus is that when hearing is damaged, the brain compensates by increasing its own internal signal gain, generating a phantom sound. Low-frequency stimulation (typically 1 Hz) is thought to suppress this hyperactivity by reducing the firing readiness of those auditory neurons.
The second target is the dorsolateral prefrontal cortex (DLPFC). The DLPFC is involved in emotional regulation and attention. Stimulating it is not meant to reduce the sound itself but to reduce how distressing and attention-capturing it feels. This is why some clinics use a dual-site protocol targeting both areas in the same session.
A typical treatment course involves 10 to 20 sessions, each lasting approximately 30 minutes, delivered over two to four weeks. Patients sit in a chair while the coil is held against their head. The sensation is often described as a tapping or clicking on the scalp. Side effects reported across trials are mild: headache and scalp discomfort are the most common, and both are transient.
The two-target rationale has an intuitive appeal. Tinnitus causes both a perception (the sound) and a response (the distress). TMS, in theory, addresses both. Whether that theory holds up in clinical trials is a separate question.
What the Evidence Actually Shows: A Plain-Language Review
What most meta-analyses agree on
Looking at the best available evidence in aggregate, rTMS does outperform sham treatment on measures of tinnitus-related distress in the short term. The two most comprehensive recent meta-analyses both support this.
He et al. (2025), which pooled data from 16 RCTs involving 1,105 chronic tinnitus patients, found that rTMS produced a mean reduction in Tinnitus Handicap Inventory (THI) scores of 11.54 points immediately after treatment, and 10.98 points at one month, compared to sham. The THI minimum clinically important difference is around 7 points, so these are real-world meaningful improvements in distress, at least in the short term.
An earlier and larger pooling by Liang et al. (2020), covering 29 RCTs with 1,228 patients, found standardised mean differences (SMDs) of 0.36 to 0.38 on distress scores at one week and one month. Effect sizes in that range are described as small-to-moderate in statistical terms, meaning the benefit is real but not large.
Where the evidence weakens
The short-term signal does not hold at six months. He et al. (2025) found no statistically significant benefit on THI at the six-month follow-up. For a condition patients typically live with for years, a treatment effect that fades within six months has limited practical value.
There is also a consistent finding across studies that rTMS does not significantly reduce tinnitus loudness. He et al. (2025) explicitly found no significant effect on Loudness Match scores (a standardised audiological test that measures how loud a patient perceives their tinnitus to be) at any time point. If you are hoping TMS will make the sound quieter, the evidence does not support that expectation. What the evidence does support, more modestly, is that the distress and interference caused by the sound may decrease for a period.
The contradictory signals
Not all meta-analyses reach the same conclusion. Dong et al. (2020), which pooled 10 RCTs involving 567 patients, found no significant improvement over sham at any time point, with a short-term SMD of just -0.04, which is essentially zero. The German S3 guideline cites this meta-analysis as one of its primary justifications for recommending against routine use (AWMF S3-Leitlinie Chronischer Tinnitus, 2022).
The largest single RCT is also a null result. Landgrebe et al. (2017), a multicentre, sham-controlled trial with 163 patients enrolled (153 completing the trial), tested 10 sessions of 1 Hz rTMS to the left temporal cortex. The adjusted mean difference in Tinnitus Questionnaire scores between real and sham stimulation was -1.0 (95% CI: -3.2 to 1.2; p=0.36), which is not statistically significant. The authors concluded that real 1-Hz rTMS over the left temporal cortex was not superior to sham, and that these findings “put efficacy of this rTMS protocol into question” (Landgrebe et al., 2017).
What comparing rTMS to other brain stimulation approaches adds
A 2024 meta-analysis by Heiland et al. (2024) compared rTMS against other neuromodulation approaches including transcutaneous electrical nerve stimulation (TENS, which uses low-level electrical current applied via skin electrodes) and transcranial direct current stimulation (tDCS, which passes a weak electrical current through the scalp) across 19 RCTs involving 1,186 patients. The finding is one of the more informative in this area: TENS and tDCS produced larger short-term reductions in THI scores (TENS: -16.2; tDCS: -19), but rTMS was the only modality to show a significant benefit in the long term, with a mean THI reduction of -8.6 (95% CI: -11.5 to -5.7) at longer follow-up.
This temporal split is worth understanding. If short-term relief is the goal, TENS or tDCS may outperform rTMS. If any sustained effect matters, rTMS has the better evidence of the approaches compared, even if that sustained effect is moderate and does not extend reliably beyond six months.
The guideline position
The German S3 clinical guideline (AWMF S3-Leitlinie Chronischer Tinnitus, 2022) reviewed all available evidence and concluded, at 92% expert consensus, that rTMS should not be used for chronic tinnitus as a routine treatment. The guideline cites both the Landgrebe null-result RCT and the Dong et al. meta-analysis showing no benefit.
A dissenting vote was filed by the German Society for Psychiatry and Psychotherapy (DGPPN), which stated that TMS “can be considered for the treatment of chronic tinnitus” in cases where other options have been exhausted, with a recommendation grade of 0 (open consideration, not a positive endorsement).
In the UK, NICE’s tinnitus guideline (NG155) does not mention TMS at all (NICE, 2020). It recommends audiological assessment, hearing aids, CBT, and sound therapy. The absence of TMS from NG155 reflects the state of UK-recognised evidence at the time it was written.
The Protocol Problem: Why There Is No Standard TMS Treatment
One reason TMS results look so inconsistent across studies is that there is no agreed treatment protocol. Published trials use stimulation frequencies ranging from 1 Hz to 20 Hz. They target the left auditory cortex, the right auditory cortex, the DLPFC, or some combination. Treatment courses range from 10 to 30 or more sessions. Some use neuronavigation (MRI-guided coil placement); most do not.
This variation means that comparing a “TMS session” at one clinic to a “TMS session” at another is not straightforward. When you read a commercial clinic’s response-rate figure, you don’t know what protocol produced it, whether it included a sham control, or whether the outcome measure had any clinical validity.
Research has not resolved this by adding complexity. A review published in 2025 found that adding DLPFC stimulation to temporal cortex stimulation has not shown superiority over temporal-only protocols, and that neuronavigation has not consistently outperformed standard coil positioning (Frontiers in Audiology and Otology, 2025). An RCT by Lehner et al. comparing single-site and triple-site stimulation found no significant difference between the two approaches.
Several trials currently recruiting are testing frequency-specific and MRI-guided neuronavigation protocols. Their results may narrow the protocol question, but that data is not yet available. Until it is, the honest answer to “which TMS protocol is best” is that nobody knows.
Who Responds Best — and Who May Not
It would be useful to predict in advance who will benefit from rTMS. The evidence here is less clear than patients or clinicians might hope.
Shorter tinnitus duration is generally associated with better outcomes, with acute tinnitus cases showing higher response rates than chronic cases. This finding is biologically plausible: the neural changes that maintain chronic tinnitus are likely more entrenched and harder to shift.
A study by Poeppl et al. (2018) examined structural brain connectivity in rTMS responders versus non-responders and found that connectivity patterns in a brain network connecting the prefrontal cortex (involved in attention and emotion), the insula, and the temporal cortex (involved in sound processing) distinguished the two groups. The clinically relevant point is that standard variables including hearing loss, tinnitus duration, and tinnitus severity did not reliably predict response. The predictor that did show some signal (brain connectivity on MRI) is not something that can be measured in a routine clinical appointment.
Comorbid hearing loss and depression are associated with poorer responses to rTMS. Patients whose tinnitus changes with jaw or neck movement (somatosensory tinnitus) may be better candidates for TENS-based approaches than for rTMS, based on mechanistic reasoning and the comparative data from Heiland et al. (2024), though a direct head-to-head trial in this specific group has not been published.
The Bottom Line: Is TMS Worth Pursuing for Tinnitus?
Here is where the evidence actually leaves you.
rTMS has a biologically plausible mechanism and a solid safety record. In most meta-analyses it reduces tinnitus-related distress more than sham treatment in the weeks after treatment ends. The short-term distress benefit appears in enough independent meta-analyses to be credible.
The limitations are real too. The effect on tinnitus loudness is not significant. Long-term benefit beyond six months is not reliably demonstrated. One major meta-analysis found no benefit at any time point. The largest single RCT found no benefit. No major clinical guideline endorses routine use: the German S3 guideline recommends against it at 92% consensus, and NICE’s tinnitus guideline does not mention it at all.
Cost is a practical barrier. TMS for tinnitus is not FDA-approved and is not typically covered by health insurance. Out-of-pocket costs range from approximately $6,000 to $15,000 for a full course.
If you have not yet fully worked through evidence-based options including CBT, sound therapy, and TRT, those are the stronger starting points: they are better supported by guidelines, more accessible, and substantially less expensive.
If you have tried those options and TMS is still on the table, the most responsible route is through a clinical trial. Trials offer protocol-controlled treatment, proper sham comparison, and often lower cost than commercial providers. Searching ClinicalTrials.gov for “rTMS tinnitus” will show currently recruiting studies.
The research is active. The protocol questions currently being studied may sharpen the picture considerably. That is not a reason to wait indefinitely, but it is a reason not to base a major financial decision on data that has yet to settle.
