The supervision contract is one of those pieces of professional infrastructure that almost every CBT supervisor knows they should have, almost every supervisor believes they do have, and almost no-one actually has in usable form. It is most often a single sheet drafted at the start of a supervisory relationship, signed in the first session, filed somewhere — usually in a folder named after the supervisee, on a laptop that may or may not still be the supervisor's working machine three years later — and never referred to again.
This is a problem in the same way that a missing fire policy is a problem. It is invisible until the day it matters, and on that day, its absence is what cascades.
What the standards actually require
The BABCP Standards of Conduct, Performance and Ethics, and the supervisor accreditation criteria that sit underneath them, both treat a written supervision agreement as a baseline requirement of competent supervisory practice. They are not unusual in this. The HCPC standards covering registered psychological professionals require documented supervision arrangements. The BPS standards for psychologists in applied roles do the same. The pattern across regulators and professional bodies is consistent — a written agreement is the expected default, not an optional extra for the meticulous.
The reason it is the expected default is straightforward. Supervision is a professional relationship with asymmetric power, clinical consequences, formative responsibilities, and confidentiality structures that diverge in important ways from those of the therapy room. The supervisor carries a duty of care toward both the supervisee and — at one remove — the supervisee's clients. The supervisee is entitled to know the terms of the relationship: what is being offered, what is being expected, what is being recorded, what gets escalated and to whom. A written contract is the mechanism by which those terms are made explicit, mutually agreed, and available for reference when something later becomes contested.
When something later becomes contested is, of course, when the contract starts to matter. Up until that point, the absence of one feels like a small administrative oversight.
What the contract is supposed to cover
The textbook frame is reasonably well-established. Proctor's three functions of supervision — normative, formative, restorative — give a structural decomposition of what supervision is for, and a contract that does not specify how each of those functions is being held is a contract that has decided to leave that work informal. Hawkins & Shohet's Seven-Eyed Supervision model offers a different cut at the same territory, focused on what the supervisor is attending to in any given moment of supervision. Both frames assume that the supervisor and supervisee have shared, explicit understanding of what kind of supervision is being delivered.
Translated into the clauses of an actual contract, this implies a document that covers:
The scope of the supervisory relationship. What modality (or modalities) is being supervised. Whether it covers a specific caseload, a service context, a training programme, or accredited practice. Whether the supervisor is the line manager or independent of line management — a distinction that affects what can be discussed in confidence and what cannot.
Frequency, duration, and modality of supervision sessions. Not just the headline "monthly, ninety minutes" but the policy on missed sessions, on rescheduling, on the minimum frequency required for accredited supervision to count toward the supervisee's CPD or accreditation evidence.
Mutual responsibilities. What the supervisee brings to each session — case material, recordings, prepared agenda items, outcomes data. What the supervisor brings — focused attention, structured feedback, willingness to challenge, availability between sessions within agreed parameters.
Record-keeping. What gets recorded in supervision, by whom, where it is stored, how long it is retained, who can access it. The supervision record is itself a clinical document; the contract is where its handling is defined.
Recording consent. If sessions or excerpts of clinical material are being recorded — for the supervisee's drilling, for the supervisor's review, for CTS-R rating, for an accreditation portfolio — the contract is where consent is established, including the supervisee's right to withdraw, the storage arrangements, and the retention period.
Confidentiality and its limits. Standard clauses on confidentiality between supervisor and supervisee, alongside the clauses that matter most — the points at which confidentiality will be overridden. Safeguarding concerns. Fitness-to-practise concerns. Concerns about client safety that the supervisor would be obliged to escalate. The contract is where the supervisee is told, in writing, what cannot be held confidentially.
Fidelity and competence monitoring. How the supervisee's competence is being tracked. Whether CTS-R review is part of the supervisory cycle. How often. With what feedback mechanism. What happens when scores fall below threshold.
Outcomes data review. Whether the supervisee's clinical outcomes data — PHQ-9 trajectories, GAD-7 trajectories, ROM data of whatever flavour the service uses — is reviewed in supervision, and at what cadence. A supervision contract that does not specify this has decided to leave outcome-based feedback to whichever weeks the conversation drifts that way.
Accreditation logging. What evidence is being captured for the supervisee's accreditation portfolio. What evidence is being captured for the supervisor's own accreditation renewal. Whose responsibility each piece is.
Complaints and resolution. What happens when supervision is not working — for either party. Escalation pathways. Mediation arrangements. The route out of the relationship.
What happens when frequency slips. The clause that almost no contract contains and that the field needs most. Supervision frequency erodes under pressure. The contract should specify what happens when a planned monthly session has not occurred for ten weeks, before it has not occurred for sixteen.
This is, by any honest reckoning, more material than fits comfortably on a single A4 page.
What contracts in routine practice actually contain
The contract that exists in most supervisory relationships, when it exists at all, covers a much smaller surface. Names. Frequency. Duration. Sometimes a note on confidentiality that gestures at the existence of limits without specifying them. Rarely anything operationally consequential about recording, fidelity, outcomes review, or accreditation evidence capture.
This is not because supervisors are negligent. It is because the template most supervisors started from was itself minimal, and because the additional clauses — the ones that matter when something goes wrong — are not part of the activation-cost-zero defaults. Writing a comprehensive contract from scratch, for each new supervisee, costs an hour or two of focused work. Multiplied across a typical supervisor's caseload of three to eight supervisees, refreshed annually, that is a non-trivial commitment to a task that has no immediate clinical reward and no immediate consequence for skipping.
The result is the contract drift that is actually visible in the field. Generic templates, signed once, never updated. Clauses that are technically present but operationally inert. Documents that would not stand up to a serious read-through by an external assessor.
Why this matters when something goes wrong
The reason this is not just an administrative grumble is that the absence of an operative contract is the structural failure from which other failures cascade. A few examples, each of which has actually happened to working supervisors in the field:
A supervisee makes a complaint about the supervision they have received — that it has been insufficient in frequency, or that the feedback has been damaging, or that the supervisor has overstepped into territory that should have been outside the supervisory remit. The supervisor's first defence, in any formal process, is the contract: the agreed terms of the relationship, the agreed frequency, the agreed scope. If the contract is a one-page document signed three years ago and never referred to since, the supervisor has very little material to work with.
A safeguarding question arises in the course of supervision and is escalated. The escalation is later questioned — was it premature, was it warranted, was the supervisee told that this would happen? The contract is where the safeguarding escalation policy should have been spelt out, with the supervisee's signed acknowledgement. Without it, the escalation looks improvised, even if it was the correct call.
An accreditation portfolio assembly begins, either for the supervisor's own renewal or for the supervisee's initial accreditation. Evidence of the supervision arrangement is required. The supervisor discovers that the contracts they have are missing, lost, stored on a previous laptop, or so generic that they would not satisfy an assessor's read-through. The wider portfolio assembly problem tends to surface this first, because it is the moment when years of informal arrangement meet a formal evidence request.
A fitness-to-practise concern emerges about a supervisee. The supervisor's role in monitoring competence is examined. The question of whether the supervisor had appropriate structures in place — whether CTS-R review was a documented part of the supervisory cycle, whether the supervisee's outcomes data was being attended to, whether the supervisor had a documented escalation pathway when concerns arose — is asked. The contract is where those structures should have been visible. If it is not visible there, the supervisor is left arguing that the structures existed informally, which is the weakest possible position.
Each of these scenarios is uncommon for any individual supervisor. Across a career, however, the probability that none of them will ever arise is not high. The contract is the piece of professional infrastructure that quietly insures against all of them, and the labour cost of building it once is small relative to the cost of not having it in the moment when it is suddenly the only document that matters.
The drift connection
There is a connection here to the wider literature on therapist drift that is worth naming explicitly. The drift literature shows that qualified therapists move away from evidence-based practice in characteristic ways when the structures that maintain fidelity are not in place. The same logic applies one level up: supervision itself drifts when the structures that maintain it are not in place, and the contract is one of those structures.
A supervisor who has no documented fidelity-monitoring clause in their contract is a supervisor whose CTS-R review cycle is whatever happens to occur in any given month. A supervisor who has no documented outcomes-data review clause is a supervisor whose attention to clinical outcomes depends on whether the supervisee mentions them. A supervisor who has no documented frequency-slip clause is a supervisor whose response to a five-week gap is whatever they happen to think on the day. None of these are failings of intent. They are failings of infrastructure, and the infrastructure starts with the contract.
The Waller and Turner observation — that drift is not driven by knowledge deficits but by the absence of structural maintenance — applies symmetrically to the supervisor's own practice. Knowing that you should review outcomes data quarterly is not what makes it happen. A contract clause that specifies the quarterly review, a calendar structure that surfaces it, a record that captures it — those are what make it happen.
What a useful contract looks like
Take all of the clauses above and a useful contract becomes a document that:
- Specifies the scope, modality, and frequency of supervision in operationally consequential terms.
- Names the recording, storage, and consent arrangements with enough specificity that an external assessor would recognise them as compliant.
- Documents the fidelity-monitoring cycle — what is reviewed, when, how, and what the threshold for action is.
- Documents the outcomes-review cycle on the same terms.
- Spells out the safeguarding and fitness-to-practise escalation pathways in language the supervisee has read and signed.
- Names what happens when supervision frequency slips, before it has slipped.
- Captures the accreditation evidence that this supervisory relationship is contributing toward, on both sides of the relationship.
- Is reviewed annually, with the review itself recorded.
A document of that scope is not a one-page template. It is a structured agreement of several pages, written in language that respects both parties' time, and — this is the part that defeats most attempts — kept current across the years of the supervisory relationship rather than treated as a one-off signing event.
The infrastructure problem
Writing a good contract once is hard but tractable. Most experienced supervisors, given a focused half-day, could produce a defensible contract for one supervisee. The work is not intellectually mysterious; the clauses are well-rehearsed in the standards literature.
The actual problem is what happens after that focused half-day. The supervisor has not one supervisee but several. Each contract needs to be specific to that supervisee's modality, training stage, accreditation pathway, and clinical context. Each contract needs to be reviewed annually, updated when accreditation requirements change, refreshed when the supervisee's role changes, and kept accessible for the supervision review process and for any future portfolio assembly. The recording-consent clause needs to track which sessions have actually been recorded and which have not. The fidelity-monitoring clause needs to connect to the actual CTS-R reviews that have been carried out. The outcomes-review clause needs to connect to the actual outcomes data that has been examined.
This is a maintenance problem, not an authoring problem. It is the same kind of maintenance problem that the accreditation portfolio assembly surfaces — a piece of professional infrastructure that one Word template cannot hold across years of practice, and that the standard approach of "I'll write it once and update it when I remember" predictably degrades into a document of historical interest only.
A live contract is a different kind of artefact. It is versioned. It is connected to the actual records of the supervisory relationship — sessions held, recordings made, CTS-R reviews completed, outcomes data reviewed. Its clauses are not aspirational descriptions of what supervision will do; they are referenced to the evidence that those things have, in fact, happened. When the supervisor or supervisee comes back to it eighteen months later, the contract does not just say what was agreed — it shows what was delivered.
That is the artefact the standards are implicitly assuming exists. In practice, it almost never does.
Supervisia's supervision platform treats the contract as a live document.
The supervision pathway carries the contract structure across modalities, with templates for clauses that the standards expect — recording consent, fidelity-monitoring cycles, outcomes-review cadence, safeguarding escalation, accreditation logging — and version-controlled updates that move with the supervisory relationship rather than freezing at signing. The contract connects to the actual evidence: which sessions occurred, which were recorded, which CTS-R reviews were carried out, which outcomes data was attended to. Portfolio assembly, complaint response, and accreditation renewal all draw from the same underlying record rather than asking the supervisor to reconstruct it.
References
- BABCP. Standards of Conduct, Performance and Ethics. British Association for Behavioural and Cognitive Psychotherapies.
- BABCP. Criteria for Accreditation as a CBT Supervisor. British Association for Behavioural and Cognitive Psychotherapies.
- HCPC. Standards of Conduct, Performance and Ethics. Health and Care Professions Council.
- Proctor, B. (1986). Supervision: a co-operative exercise in accountability. In M. Marken & M. Payne (Eds.), Enabling and Ensuring: Supervision in Practice. National Youth Bureau.
- Hawkins, P. & Shohet, R. (2012). Supervision in the Helping Professions (4th ed.). Open University Press.
- Waller, G. & Turner, H. (2016). Therapist drift redux: Why well-meaning clinicians fail to deliver evidence-based therapy, and how to get back on track. Behaviour Research and Therapy, 77, 129–137. DOI: 10.1016/j.brat.2016.01.007. PubMed: 26752326.
Last updated: May 2026