Art Therapy for Children: Schools, CAMHS and Who the Picture Belongs To

Art therapy with a child is a clinical intervention in which making an image is part of a therapeutic relationship, delivered by a practitioner holding a title protected in UK law, and the questions that separate it from every other creative hour in a child’s week are who agreed to it and what happens to the picture. Most pages on this subject describe children expressing feelings they cannot put into words. That part is real, and it is also the easy part. The difficult and largely unwritten part is that the client is eight, cannot lawfully consent, may be in the care of a local authority, and has just made something that will be filed rather than pinned up.

I do not work with children, I have never sat in an art therapy session of any kind, and I am not going to write as though I have. What I can offer is adjacent and slightly uncomfortable. A primary school head asked me in to run singing for one class, which she described in the first thirty seconds of the call as a difficult class. I asked what she wanted to be different afterwards, and she said, more or less, that they would be calmer for the rest of the morning. I said I would happily come and sing with them, that they would probably enjoy it, that I could promise nothing about the rest of the morning, and that if the difficulty was really one or two children then singing at all thirty was an odd instrument to reach for. What I noticed, and have thought about since, is that the school was not asking for something for the children. It was asking for something to be done about them, and the two requests wear very similar clothes. The rest of this page comes from the profession’s published standards, from the reviewer who checks it, and from what services say about how the work is arranged.

Claims about effect on this site carry a tier: Tier 1 supported, Tier 2 promising but limited, Tier 3 practice and anecdote. What each label requires, and why an empty one is not a verdict, is on reading arts in health research.

Where children are actually seen

Six settings account for most art therapy with children in the UK, and none of them rests on a national entitlement.

Schools come first by volume, mainstream and special, sometimes with a therapist employed directly, more often through a commissioned block bought with a fixed pot of money. Child and adolescent mental health services employ arts therapists in some areas and have never had one in others. Services for looked after children use arts therapies heavily, partly because the referrals often involve early experiences that are difficult to talk about directly and partly because those services have their own budgets. Paediatric wards and children’s hospices host therapists where the funding exists, alongside a separate hospital arts programme that is not therapy, covered on arts in hospitals and creative work at the end of life. And the voluntary sector runs a great deal of it, project by project.

The practical consequence is that two children with identical needs a few miles apart can have completely different access, and that a course of work is often as long as a grant rather than as long as the child needs. Ask at the start how many sessions are funded and what happens afterwards, because a planned ending is part of the clinical work and an ending imposed by a spreadsheet is not.

Why the modality is used with children, and the limit of that reason

The usual rationale is that making an image places a lower demand on verbal expression than talking does, which suits children whose vocabulary for their own internal state is still forming, or who have reasons not to say things out loud.

That is a coherent reason to try the approach and it is not an evidence claim, and the distinction is worth holding on to, because this is exactly the point where rationales get promoted. A plausible mechanism explains why somebody thought this would work. It does not establish that it does.

Services reach for it for a second, blunter reason: a child will often stay in a room with materials considerably longer than in a room with a question. Whether that is therapeutic benefit or simply better attendance is the sort of thing a controlled study would have to separate. What happens in an art therapy session describes the shape of the hour, and art therapy against an art class draws the line that most school newsletters do not.

Who consents, and who holds parental responsibility

Consent for a child’s therapy normally comes from whoever holds parental responsibility, and the first useful move is to establish who that is rather than assuming it is the adult in the room.

Parental responsibility can sit with one parent, with both, with a step parent or guardian who has acquired it, and with other adults through a court order. Where a child is subject to a care order, the local authority holds parental responsibility alongside the parents, which changes the conversation entirely: the decision maker may be a social worker, the reporting may run into statutory review meetings, and the therapist may be one voice among many in a process with its own timetable. For children accommodated by an authority without a care order, the position is different again, and parents remain the consenting party. None of this is exotic. It is the ordinary situation for a large share of the children referred to arts therapies.

Alongside legal consent sits the child’s own agreement, which is not the same thing. A young person judged competent to understand what is proposed can consent for themselves, a principle usually named after the Gillick case, and older teenagers are generally treated as decision makers in their own right. Below that, what matters is assent: whether the child, in whatever way the child communicates, is willing. A child brought under protest can be made to sit in a room and cannot be made to use it, and a therapist who treats reluctance as an obstacle rather than as information is not doing the work properly.

The general framework, including capacity, recording and photography, is on consent and safeguarding in arts in health, and the parallel questions for a modality where the goal itself is contested are on music therapy for autistic children.

What happens to the work a child makes

This is the part almost nobody explains in advance, and it is the reason this page exists. The images a child makes in art therapy are part of the clinical record. They are not fridge pictures.

In practice that means the work is normally stored by the service, with the notes, under the same confidentiality and retention arrangements as any other clinical material. It is not displayed. It is not sent home in a book bag. It is not shown to the class. It may be looked at again in a later session, with the child, as part of the work. Who can access it is governed by the service’s records policy rather than by who feels entitled to see it, and that policy is a fair thing to ask for in writing.

Three situations complicate it, and all three are common:

• A child wants to take something home. Many therapists will allow it for a particular piece, and some will not, and the reasoning is clinical rather than administrative: an image that leaves the room enters a household where it will be interpreted, praised, questioned or thrown away, and none of that is neutral for the work.
• A parent wants to see the work. Parents holding parental responsibility have rights of access to information about their child, and therapists commonly meet this by sharing themes and progress rather than by handing over images. If the child has made something about the household, the tension there is obvious and should be named early rather than negotiated in a crisis.
• A report is written for a review meeting, a tribunal or a court. This is where the record stops being private in any ordinary sense. A therapist may be asked to describe the work, and occasionally to produce or characterise particular images, in proceedings whose purpose is not therapeutic. Children are entitled to be told at the outset that this can happen, in language they can follow, and services that skip that conversation are storing up a betrayal.

Ask, before anything starts: where is the work kept, who can see it, does anything go home, is it returned at the end or destroyed, and what happens to it if we move away or the school stops commissioning.

Confidentiality, and the limit the child must be told about

What is said and made in the room is treated as confidential, with one exception that is not the therapist’s to waive, and the child should hear about it at the beginning rather than at the moment it is used.

The exception is risk of significant harm, to this child or to another. If something emerges that meets that threshold, it is passed on, and the therapist’s discretion extends to how and when the child is told rather than to whether the disclosure happens. Good practice is to explain this in the first session in words a child can actually process, something closer to “if you tell me something that makes me think you or another child are not safe, I have to tell someone who can help, and I will tell you when I am doing it” than to a paragraph about safeguarding policy.

Below that threshold, what leaves the room is usually a summary: how the child is using the sessions, what is changing, what would help elsewhere. School staff frequently expect more, and a therapist declining to provide a running commentary is not being obstructive. It is the condition that makes the room worth having.

The school setting, and the referral that is set up to fail

School based work has its own physical and social problems, and they are not trivial: the room, the timing, and what the referral is actually for.

The room is often whatever is free: the medical room, a corner of the library, or a space with a glass panel in the door that classmates walk past twice an hour. Confidentiality in a school is thin by construction, because everyone notices who leaves the classroom and when. Timing is the second problem: a session scheduled against a subject the child loves creates a cost for attending, and one scheduled against a subject the child dreads creates an incentive that has nothing to do with the work. Weekly consistency is harder to protect here than anywhere, because assemblies, trips, tests and staff absence all outrank it.

Then the referral itself. The most common way this work is set up to fail is the fix this child referral: a child whose behaviour is difficult for adults is sent out of the classroom to be repaired, while nothing in the classroom, the timetable, the household or the adult responses changes. The therapist is then holding work whose implicit goal is compliance, with a child who has correctly understood that they are the identified problem. Sometimes the difficulty really does sit with the child and the referral is right. Often it does not, and a good service will say so, which is an awkward conversation with a school that has paid for a block of sessions.

The question that opens this up is simple: what else is changing alongside this. If the honest answer is nothing, the therapy is being asked to do a job that is not therapy’s.

The evidence position, and what would have to exist

No effect size for art therapy with children appears on this page, and that is a decision rather than an oversight.

No result in this area has been checked against its source document by the reviewer who signs off figures on this site. That is the entire reason, and it is a statement about verification here rather than a description of the literature. Tier 3, of the nobody has verified it here kind, not the null kind. Those two situations get labelled identically in most writing about this field and they are not the same. Where the controlled evidence exists and is genuinely null, this site says null in those words, and it has somewhere to point when it does.

The place it points to is MATISSE, worth knowing about precisely because it is not about children. That trial randomised 417 people with established schizophrenia across four UK centres: 140 to group art therapy, 140 to activity groups as an active comparator and 137 to standard care alone, with 355 (85%) followed up at 2 years¹. Against standard care at 24 months it reports global functioning at an adjusted mean difference of -0.9 (95% CI -3.8 to 2.1) and symptoms at 0.7 (95% CI -3.1 to 4.6), with no difference between the three arms on either primary outcome and no secondary outcome favouring art therapy². Tier 3, of the null controlled evidence kind, not the nobody has looked kind. Two things belong in the same breath: only 86 of the 140 randomised to art therapy (61%) attended even one group, which makes it a test of referral rather than of therapy received; and it tested weekly group work with adults who had established schizophrenia on those two outcomes, so it says nothing about a nine year old in a school room. It also averages across hundreds of people and predicts nothing about any one of them. The detail is on art therapy for psychosis, and the reading habits behind it on reading arts in health research.

What would have to exist for a number to appear here: a controlled trial in children, adequately powered, with the comparator named, the primary outcome named in advance and measured by someone blind to allocation, the sample size stated, an effect size with its confidence interval, a certainty rating, follow up beyond the end of the sessions, and a year. Or a systematic review of such trials. Nothing less gets printed.

And the standing caution in the other direction. The absence of a verified figure is not a finding that this does not help. Arts therapies with children are hard to blind, hard to randomise, funded in short cycles, and aimed at outcomes that are difficult to measure well. Nor does the size of the literature settle it: the WHO Europe scoping review mapped over 900 publications, made up of 200 plus reviews and 700 plus individual studies, with those reviews between them covering over 3,000 studies³. Those counts nest rather than adding up, and a scoping review maps a field instead of pooling it, so it carries no tier at all.

The register check, and what to ask

Art therapist and art psychotherapist are two separate titles protected in UK law, covering one profession. With dramatherapist and music therapist that makes four protected arts therapy titles, and the sector’s own habit of writing about “the three arts therapies” quietly loses one of them.

The HCPC register listed 6,103 arts therapists across all four titles in July 2026⁴, and using a protected title without registration is a criminal offence. Searching a practitioner’s name takes a minute and is the most informative thing you can do before a first session. The professional body for the visual side maintains a members directory and publishes practice standards⁵; membership is not the same as registration and both are worth checking. If the person is not registered, that is a fact about what they may lawfully call their work rather than a judgement about it. Training as an art therapist sets out how somebody gets onto that register, dramatherapy covers a fourth title often offered to the same children, and therapeutic writing and photography covers approaches outside the register entirely.

Questions worth putting to a school or a child and adolescent mental health team, in roughly this order: is the practitioner on the HCPC register; who is consenting and who else has to be told; how many sessions are funded and what happens then; where will the sessions be and what will the child miss; what is written down and who reads it; what happens to the artwork; what would you tell me if it were not working; and what changes elsewhere while this goes on.

Every one of those has a dull, factual answer, and a service that finds them intrusive has told you something. None of it settles whether art therapy will help a particular child, which nobody can honestly promise. And none of it is a reason to alter anything else in a child’s care, because this work is arranged alongside education, medicine and family support rather than in place of any of them. The pillar page for the whole profession is art therapy.