Music Therapy: What It Is, Who Delivers It, and What the Trials Show

Music therapy is a clinical intervention delivered by a registered practitioner who uses live music to work towards assessed, agreed goals, and in the United Kingdom the title is protected in law, which makes it a different thing from a singing group even when the two look similar from the corridor. The evidence behind it is real, uneven, and much more specific than the claims made for it: strong for gait speed after stroke, strong for anxiety before surgery, small for depressive symptoms in dementia, and pointing the other way for agitation in dementia.

I am on the other side of this line. I run weekly singing and songwriting sessions in two care homes and on a stroke rehabilitation unit, as a community arts practitioner, and I am not registered to call any of it therapy. The first time I sat in on a registered music therapist’s session I expected a better version of what I do. It was not that. It was quieter, slower, mostly one to one, and organised around a goal written down in advance that had nothing to do with whether anybody enjoyed the hour. Both things are worth having. Only one of them is a clinical intervention, and the distinction is set out in full on music therapy against community music.

Every efficacy figure below carries a tier: Tier 1 supported, Tier 2 promising but limited, Tier 3 practice and anecdote. What the labels mean and how to check them is on reading arts in health research.

Who is allowed to call themselves a music therapist

In the UK, only a practitioner registered with the Health and Care Professions Council, and using the title without registration is a criminal offence rather than a matter of professional etiquette.

Music therapist is one of four protected arts therapy titles: art psychotherapist, art therapist, dramatherapist and music therapist. You will often see “the three arts therapies” in sector material, and it is wrong; art therapist and art psychotherapist are two separate protected titles covering the same profession. In July 2026 the HCPC register listed 6,103 arts therapists in total across all four titles¹, which is a small profession by any measure and explains a great deal about waiting lists.

Getting on that register means a postgraduate qualification, normally a masters with supervised clinical placements, and it is the route described on training as a music therapist. Outside the UK the picture changes completely. In the United States the credential is MT-BC, awarded by the Certification Board for Music Therapists, and there were 10,699 MT-BC credential holders as at 30 June 2026 according to CBMT’s certificant data². Worth knowing if you go looking: the board’s own About page still describes the credential as held by “over 9,000” practitioners, which is stale, so use the certificant data and check its date. Most US states do not license music therapists at all, which makes the credential the only formal check available. How to find a music therapist covers the registers country by country.

What actually happens in a session

A session is built around a goal that was agreed before it started, and the music is the means rather than the point.

In practice that means live music far more often than recorded, improvisation more often than performance, and a great deal of the therapist following rather than leading. Instruments are usually chosen for being playable without training: percussion, tuned bars, a piano or guitar the therapist plays. Nobody is assessed on their playing and nothing has to sound good. Around the sessions sits the unglamorous half of the job that distinguishes it from an activity: an initial assessment, written goals, session notes, clinical supervision, multidisciplinary meetings, and a planned discharge conversation rather than a course that simply stops. What happens in a music therapy session goes through it in detail, and anything happening in a ward or a care home also runs into consent and safeguarding, which is a harder problem here than people expect.

Where the evidence is strongest: walking after stroke

The best evidenced claim in the whole of arts in health is not about mood at all. It is about gait speed, and it comes from rhythmic auditory stimulation.

The Cochrane review of music interventions after acquired brain injury reports gait velocity 11.34 m/min faster with rhythmic auditory stimulation (95% CI 8.40 to 14.28), from 9 trials and 268 participants, rated moderate quality³. That is Tier 1, and it is the one place on this site where the label is unambiguous.

Two things about it are routinely lost. It is reported in metres per minute, and converting it into another unit produces a livelier looking number and no new information, so this site does not. And it is a claim about walking speed, full stop. It says nothing about mood, communication, confidence or quality of life after stroke, and a page that let it drift into those would be misusing the strongest evidence in the field. Rhythmic auditory stimulation covers what the trials measured and who it suits; music therapy after stroke covers the wider picture, where the evidence is much thinner.

Anxiety before surgery

Music before an operation reduces anxiety, and the review behind that claim is unusually clean for a soft outcome.

Twenty six trials, 2,051 participants, anxiety 5.72 STAI-S units lower (95% CI -7.27 to -4.17)⁴. Tier 1, with one honest caveat: the review was published in 2013, before certainty assessment became routine, so it carries no certainty rating and this site will not attach one to it. What it has instead is a decent trial count, a decent sample and a tight interval, which is more than most claims in this field can show. See music before surgery and anxiety.

Note the intervention here is often recorded music offered around a procedure rather than a course of therapy with a registered therapist, which is a reason to be careful about crediting the profession with the finding. How programmes of this kind are organised inside a health building, and who runs them, is on arts in hospitals.

Dementia: a small real effect, and a null one that matters more

In dementia the current evidence supports a small effect on depressive symptoms and does not support the effect on agitation that is claimed most often.

The 2025 Cochrane review of music based therapeutic interventions covers 30 studies and 1,720 participants randomised, and reports⁵:

Outcome	Result	Certainty	Tier
Depressive symptoms	SMD -0.23 (95% CI -0.42 to -0.04)	Moderate	Tier 1
Agitation and aggression	SMD -0.05 (95% CI -0.27 to 0.17)	Moderate	Tier 1, null result
Anything persisting 4 weeks after treatment ends	No evidence of it		Tier 3

The agitation row is the one to sit with. A narrow interval centred on zero, at moderate certainty, is a finding that music based interventions do not meaningfully change agitation. It is not a gap waiting to be filled by a better study, and it is close to the opposite of what most secondary coverage says this review found. Anything citing 22 studies and 1,097 participants is quoting the superseded 2018 version.

The review contains no separate quality of life estimate, so any quality of life number attributed to it has come from elsewhere. For scale: 57 million people worldwide were living with dementia in 2021⁶, which is why a small effect on one outcome still matters and why the claims attached to it are so hard to dislodge. Music therapy for dementia goes through the review in full, and creative activity in dementia care covers what good practice in a home actually looks like. Two things frequently mistaken for music therapy in dementia have their own pages, because their evidence is separate and so is the way they are arranged: personalised playlists, which is a headphones intervention rather than a therapy, and dementia choirs and singing groups, which is community activity of the kind I run myself.

What NICE guidance does and does not say

NICE does not recommend music therapy for dementia. A full text search of NG97, the 45 page guideline on dementia assessment, management and support, returns zero occurrences of music, art, dance, drama or creative activity⁷.

I want to be careful about what that does and does not establish. NG97 does address non-pharmacological approaches to distress and agitation in general terms, and it does not tell anybody to stop offering music. What it does not do is name any arts intervention, in any recommendation, anywhere. So “NICE recommends music therapy” is not a strong claim stated loosely; it is a claim about a document that the document does not contain. It circulates because it is repeated by people quoting other people who were quoting a summary, which is exactly the failure mode described on reading arts in health research.

Cancer care: a lot of trials, very little certainty

In cancer care the pooled evidence looks impressive by volume and is rated very low certainty, which is the field’s best lesson in why those are different things.

The 2021 Cochrane review covers 81 trials and 5,576 participants and reports anxiety 7.73 STAI-S units lower and pain at SMD -0.67, both at very low certainty⁸. Tier 2 for both. Eighty one trials is a large literature; very low certainty means the reviewers could not trust the pooled answer, chiefly because the individual trials were small, unblinded and inconsistent. Both facts are true at once. See also art therapy during cancer treatment, where the certainty problem is the same shape.

Where the evidence is not settled

For several of the uses music therapy is best known for, this page quotes no effect size, and that is deliberate rather than an omission.

Music therapy with autistic children, in palliative care, and in adult depression and anxiety are all areas where trials exist and where the outcome measures, the comparators or the certainty ratings need setting out properly rather than summarising in a sentence. Those pages are music therapy for autistic children, music therapy in palliative care and music therapy for depression and anxiety. A figure quoted here without its design, interval and certainty would be exactly the kind of citation this site exists to stop repeating.

And the standing caution: the absence of a quoted number is not a verdict. Tier 3 means the question is open, not closed against. A great deal of music work in care settings has never been studied properly for reasons that have nothing to do with whether it helps.

What it costs and how people get it

Most people who receive music therapy do not pay for it directly, because it reaches them through a service rather than a private booking.

In the UK, the British Association for Music Therapy recommends from £52.00 upwards for an individual session in private practice⁹. That is a recommended floor, which is what it is; it is not an average and not a typical price, and actual fees vary by region, setting and whether the work is one to one or in a group. In the United States, the 2021 AMTA workforce analysis reports a median of $79 per hour¹⁰. Quote the median rather than the mean, which is skewed upwards by outliers, and date it 2021, because it is not a current figure.

Where the work comes through the NHS, a hospice, a school, a local authority or a charity, it is normally free to the person receiving it and the real constraint is whether a post exists in your area. With 6,103 arts therapists on the register across all four titles¹, for populations in the tens of millions, availability is the binding limit far more often than money is. Practical routes are on how to find a music therapist, and for the community side, arts on prescription and how to get referred to an arts scheme.

Nothing here is a reason to change anything about anyone’s treatment, and no arts intervention substitutes for one. If someone offers you music therapy and is not on the register, the useful question is not whether they are any good. It is what their training was, and what they are calling the thing they do.