Physical therapy can reduce tinnitus linked to neck problems, at least in the short term: in a randomised trial of 38 people, 53% improved substantially after six weeks of treatment, but only 24% kept that improvement six weeks later.
That applies to one kind of tinnitus. Somatic tinnitus is tinnitus that the neck or jaw can modulate, and TMJ disorder and tinnitus covers the jaw version of that pathway. This article looks at what happens when the neck and jaw are treated directly through structured physical therapy, which exercises were tested, and what the trials found.
What the treatment involves
Studied physical therapy for cervicogenic somatic tinnitus (CST) combines hands-on treatment with exercises the patient continues at home. The hands-on part is usually manual mobilisation of the cervical spine, and in some trials stretching and techniques aimed at myofascial trigger points and muscle tension in the head and neck. It is a different intervention from chiropractic spinal manipulation: closer to what a physiotherapist would do for chronic neck pain, applied to people whose tinnitus is believed to have a neck or jaw component.
The neck and jaw exercises that were studied
There is no single standard routine behind the research. Each trial below tested a package, in most of them the therapists were free to adapt the exercises to the patient, and one small pilot used a fixed programme. What the packages contained is still the most useful guide to what "exercises for tinnitus" means in the research.
Deep neck flexor and self-mobilising exercises. The trial behind this article's headline numbers gave up to 12 sessions of 30 minutes over six weeks. Its published protocol describes manual mobilisations, exercise therapy, and a home exercise booklet adapted for tinnitus patients, including craniocervical flexion exercises for the deep neck flexor muscles, done lying down and sitting, and self-mobilising exercises. The therapists could adapt the techniques and exercises to the patient's situation at each visit.
Neck exercises alone, by app. A 2024 pilot trial tested exercises without a therapist's hands. Thirty-eight adults with chronic somatic tinnitus were randomised to a nine-week app programme or to no treatment. The programme worked on strength, endurance and coordination of the deep neck flexor and extensor muscles and the shoulder-stabilising muscles, plus mobility and posture exercises: 16 exercises at five difficulty levels, with three exercises each week done daily from video instructions.
Among the 31 who finished, the app group's Tinnitus Handicap Inventory score fell from an average of 51.3 to 38.9, while the untreated group's went from 49.8 to 53.0. That is a promising start and no more. The control group received nothing rather than a sham, there was no follow-up beyond nine weeks, the authors call it a pilot to be interpreted with caution, and one author declares a stake in a company developing e-health applications.
Stretching with trigger-point work. A 2025 German trial compared treatment against a waiting list in 70 people whose tinnitus or dizziness changed when head and neck muscles were pressed. The treated group received manual therapy and trigger-point treatment, plus stretching exercises that the therapist demonstrated and patients did at home for about 15 minutes a day. They averaged 17 sessions of 20 to 30 minutes, spread over 9 to 28 weeks. Eighteen of the 37 treated patients improved by at least 17 points on the Tinnitus Handicap Inventory, against 2 of the 33 on the waiting list. The comparison was a waiting list, the same physician assessed everyone, and outcomes were measured only one to two weeks after treatment ended.
Jaw exercises and habit reversal. A 2020 Belgian trial treated 80 people whose tinnitus was linked to the jaw joint. For up to 18 sessions over nine weeks, they received massage of the chewing muscles, stretching exercises, relaxation therapy, and counselling on mouth habits, teeth grinding, sleep hygiene and lifestyle, with biofeedback. People who ground their teeth also got a splint from a dentist — 52% of participants — and neck treatment was added where needed.
The result was split. On the trial's primary measure, the Tinnitus Questionnaire, the average change did not differ significantly between groups: scores fell 4.1 points against 0.2. More treated patients did move down a severity grade on that questionnaire, 30.0% against 2.8%, which was significant. On the Tinnitus Functional Index, scores fell 13.8 points against 5.0, also a significant difference.
Jaw and neck exercises together. A 2020 Spanish trial gave all 61 of its patients, whose tinnitus was attributed to a jaw disorder, six sessions over a month of cranio-cervical and jaw exercises, self-massage and education. One randomised group also received hands-on treatment of the jaw joint, neck and chewing muscles, and that group did better on tinnitus severity and the Tinnitus Handicap Inventory, with a blinded assessor following them for six months. A 2026 trial of 31 people had the same shape: everyone did daily home exercises, and the group that added six weekly manual therapy sessions improved more.
Notice what those last two designs can and cannot say. Both show that hands-on treatment added something on top of exercises. Neither tells you how much the exercises did on their own, because everyone got them.
What the main neck trial found
The best-known trial is a delayed-start randomised controlled trial of 38 patients with cervicogenic somatic tinnitus and neck complaints, comparing an immediate-treatment group against a group receiving the same protocol six weeks later. It found substantial short-term results: 53% of patients experienced substantial improvement in their tinnitus immediately after the 6-week treatment course. Neck pain and function scores also improved significantly and durably. The tinnitus benefit, however, was considerably less durable than the neck benefit: only 24% of patients maintained that substantial tinnitus improvement at the 6-week follow-up after treatment ended.
Why that drop-off matters
Physical therapy for CST appears to produce a real, measurable short-term effect on tinnitus for around half of patients, but that effect faded for roughly half of those responders within six weeks of stopping treatment. The neck pain and function improvements held up much better over the same period. The most natural reading is that the treatment reliably helps the neck problem, and that the tinnitus connected to it may need ongoing treatment, such as continuing the home exercises or maintenance sessions, to stay improved.
Who this is for
This approach is specifically relevant if your tinnitus fits the somatic pattern: changing in loudness or pitch with neck movement, jaw clenching, or pressure on specific neck and shoulder points. Without that pattern there is no clear mechanistic reason neck or jaw treatment would be expected to help, and it should not be assumed to be a general tinnitus treatment.
The 2025 trial's authors add a condition worth keeping: manual therapy may help cervicogenic somatosensory tinnitus provided other potential causes, including ear, nose and throat conditions, have been ruled out. A new tinnitus deserves an assessment before a treatment plan.
A 2026 review looked for the tinnitus effect and did not find it
The trials above are the strongest evidence here. A 2026 systematic review in BMC Musculoskeletal Disorders took a wider view — physiotherapy for neck pain and the symptoms that travel with it, including cervicogenic dizziness and tinnitus — and came back with much less.
It analysed 13 studies covering 785 patients. Of those, 10 were at high risk of bias, one raised some concerns, and two were at low risk — so 77% of the literature it found was of poor methodological quality, which the review states as its headline conclusion.
What it could support: Mulligan mobilisation produced greater improvement in cervical range of motion than placebo treatment, on moderate evidence, with no difference in balance. Effects on pain intensity and dizziness-related handicap were conflicting.
What it could not: limited or no evidence for tinnitus-related factors, alongside pain, dizziness and psychological measures and objective functional status — whether physiotherapy was compared against a waitlist, against placebo, or against a different form of physiotherapy.
That is a real check on this article rather than a footnote to it. A treatment can have promising trials in a well-selected group — people with demonstrable somatic tinnitus — and still have no pooled evidence across the broader population of people with neck pain and tinnitus. Both of those are true here, and they are not in conflict: the trials above selected their patients carefully, and this review did not.
The practical consequence is about selection. If your tinnitus changes when you move your neck or jaw, the trial evidence above is about people like you. If it does not, this review is the better guide to what physiotherapy for a sore neck is likely to do for the sound, which is probably nothing.
Finding the right help
The 2024 VA/DoD tinnitus guideline suggests a multidisciplinary approach to assessing and treating bothersome tinnitus that comes with a jaw disorder, neck dysfunction or both — a weak recommendation, which is the strongest any tinnitus treatment received from it. Its provider summary lists the people involved: audiologists, dental providers, physical therapists, orofacial massage providers and otolaryngologists.
In practice that means an audiologist or ENT assessment first, then a physiotherapist experienced with cervicogenic or jaw-related tinnitus rather than general spinal manipulation, and a dentist if clenching or grinding is part of the picture. Go in expecting the neck or jaw to improve more reliably than the sound, and treat the home exercises as something to keep doing rather than a course to finish.