Sand Tray Therapy Documentation Examples for Progress Notes

How to Document Sand Tray Therapy Without Over-Interpreting the Scene

Sand tray work is symbolic and largely nonverbal, but the chart is concrete and defensible. That tension is what makes the note hard to write. There are two failure modes. The first is over-interpretation — "the wall means avoidance," "the dragon means anger" — which assigns a fixed meaning the client never gave and reads as diagnosis-by-symbol under audit. The second is under-documentation — "did sand tray, client engaged" — which says nothing about what happened, what you did, or why it was clinically necessary. This post is the deep documentation reference for both problems: full worked SOAP and DAP examples for the same session, a reusable framework, and the privacy rules for attaching the scene image.

After a session in the free virtual sand tray, the next task is the note — that is what this post is about. If you need the clinical frame instead (telehealth setup, prompts for children and teens, and safety boundaries), the online sand tray telehealth guide covers that, so this post stays focused on documentation. This is clinician education, not patient treatment advice; the examples below are all fictional and de-identified.

What Actually Belongs in the Note (Describe the Build, Not the Symbol)

Sandtray documentation guidance is consistent: record observable behavior and actions concretely, and state what you did with them. Southern Sandtray Institute's documentation guidelines advise noting whether the client was silent or hyper-talkative, how they handled the sand and the miniatures, which miniatures were placed, and any changes or revisions made during the session — and whether the therapist reflected, processed, tracked, or summarized the client's movement in the tray (Southern Sandtray Institute, "5 Guidelines to Documenting Sandtray Therapy Sessions"). A practical observation checklist:

• Engagement and pace — silent, talkative, hesitant, absorbed; how the client handled the sand.
• Miniatures — roughly how many, and which ones were placed (briefly, by category).
• Placement and structure — distance, grouping, barriers, what sits at the center, what sits at the edge.
• Movement and change — what was added, removed, or revised during the session.
• Client language — the client's own words, quoted or paraphrased, with no identifying details.
• Affect and regulation shifts — engagement, avoidance, activation, or settling across the build.
• Therapist stance — whether you reflected, tracked, processed, or summarized.

The governing rule: stay objective and focus on what was observed rather than interpreting the client's actions or emotions (Mentalyc, "Play Therapy Progress Notes"; Seven Stones Mental Health, "Documenting Play Therapy Notes"). Attribute meaning to client report, not to the symbol.

The Golden Thread: Why the Note Has to Connect to the Plan

A descriptive note is necessary but not sufficient. A play therapy progress note may be reviewed for insurance claims, billing, court cases, or audits, so it should contain language that supports the service provided (Mentalyc). That support comes from the "golden thread" — the consistent clinical connection running through the record: the intake assessment identifies the problem, the treatment plan sets goals for that problem, and each progress note shows that the services delivered match what the plan prescribed. This alignment is how clinicians provide clear evidence of medical necessity (TherapyNotes, "The Golden Thread"; SonderMind, "Using Medical Record Documentation Templates"). It is reinforced by reviewing the treatment plan on an ongoing basis and by including objective measures that assess baseline function and progress rather than asserting it (PacificSource, "Behavioral Health treatment reviews"). Exact review cadence is payer- and level-of-care-dependent, so check your contracts rather than assuming a fixed interval.

For expressive work, the golden thread is what separates a defensible note from "play for play's sake." A scene that reduced an anxious child's distress is documentable as an emotion-regulation intervention tied to a plan goal; the same scene with no goal link reads as recreation to a payer. Where a measure is the anchor, name it — for example, a PHQ-9 re-administered at set intervals — so progress is measured rather than asserted.

A Reusable Sand Tray Note Framework

You do not need a sand-tray-specific note template — you need to map the observation checklist above onto whatever note sections you already use. The same content slots into SOAP, DAP, and BIRP differently:

• SOAP has four sections — Subjective (the client's report: reason, words, history), Objective (your concrete observations of the build), Assessment (clinical reasoning tied to the goal), Plan (next step). Put client language in S, the build in O.
• DAP has three sections and combines the subjective and objective information into one Data section, then Assessment and Plan. Because it has one section fewer, DAP is regarded as brief and faster to write, which is why clinicians with large caseloads often favor it for routine documentation (Upheal, "SOAP vs. DAP notes"). The same client words and the same build live together in Data.
• BIRP has four parts — Behavior, Intervention, Response, Plan — and foregrounds the technique used and the client's reaction to it, which keeps the clinician's methods and the client's response clearly separated and is widely used in behavioral health to support medical necessity, reimbursement, and audit readiness (ICANotes, "BIRP Notes"). It fits sand tray well because the sand tray is the intervention.

Pick one format and stay consistent across the chart. The examples below show the identical session written two ways, so you can see the observations map cleanly between formats.

Example 1 — SOAP: Open-Ended Adult Sand Tray Session

A fictional, non-PHI SOAP note from an open-ended adult session. The treatment-plan goal in this composite is reducing avoidance and building one usable grounding strategy.

S (Subjective): Adult client stated they "wanted to show how the week felt instead of talking about it" and described feeling "stuck between wanting help and not trusting it." Reported elevated stress at work and difficulty asking for support.

O (Objective): Client built an open-ended scene using roughly eight miniatures: a central figure, a small cluster of figures set at a distance, one protective object near the edge, and a barrier between the central figure and the cluster. Client worked deliberately and narrated as they placed items, describing the central figure as "wanting to be close but not safe yet." Affect was engaged and slightly tearful while describing the barrier; client settled when identifying the protective object. Therapist tracked placements and reflected the client's language without assigning symbolic meaning.

A (Assessment): Scene and narration are consistent with the plan target of approach–avoidance around support-seeking. Client demonstrated capacity to name a felt distance and to identify a self-described source of safety, which supports continued work on the avoidance goal. Meaning anchored to client report; no fixed interpretation of objects made.

P (Plan): Continue expressive work toward the avoidance goal. Client identified one grounding strategy to rehearse before next session. Re-administer the PHQ-9 at the next scheduled interval; revisit support-seeking in the following session.

What makes this note defensible: it describes the build concretely, quotes the client rather than decoding the objects, and ties the intervention to a named plan goal and measure.

Example 2 — DAP: Same Session, Different Format (Telehealth Virtual Sand Tray)

The same fictional session, written in DAP, and conducted as a telehealth screen share using the virtual sand tray. DAP folds the subjective and objective information above into one Data section.

D (Data): Telehealth session conducted by screen share using an online virtual sand tray. Adult client stated they "wanted to show how the week felt instead of talking about it" and described feeling "stuck between wanting help and not trusting it." Client built an open-ended scene with roughly eight miniatures — a central figure, a cluster of figures set at a distance, a protective object near the edge, and a barrier between the central figure and the cluster — narrating that the central figure was "wanting to be close but not safe yet." Affect was engaged and slightly tearful at the barrier and settled when the protective object was placed. Therapist tracked placements and reflected client language.

A (Assessment): Presentation is consistent with the plan target of approach–avoidance around support-seeking; client named a felt distance and a self-identified source of safety. Meaning anchored to client report.

P (Plan): Continue expressive work toward the avoidance goal; client to rehearse one grounding strategy before next session; re-administer PHQ-9 at the next interval.

Demonstrating the same observations in two formats is the value here — most templates give one format and stop. The format is a container; the clinical content does not change.

Example 3 — Child/Teen Regulation or Coping Session

A broad, fictional, non-PHI child/teen example with a regulation/coping focus. Note the developmental adjustment to language — concrete and plain, not interpretive — and a single caregiver-context line.

B (Behavior): Pre-teen client engaged readily, talkative throughout, and built a scene with roughly six miniatures: a small figure, two "helper" animals the client named, and a fence the client said was "to keep the loud stuff out." Client described feeling "wound up after school." Caregiver context noted at check-in: increased after-school irritability per caregiver report.

I (Intervention): Therapist used the sand tray as an emotion-identification and coping intervention, tracked the build, reflected the client's words, and supported the client in naming one calming strategy connected to the "helper" figures.

R (Response): Client identified two body cues of being "wound up," rehearsed a paced-breathing strategy in session, and reported feeling "more calm" by session end. Engagement remained high; no distress observed.

P (Plan): Practice the named calming strategy after school; review with caregiver at next check-in; continue coping-skill goal next session.

This stays broad and non-prescriptive — it documents the build and the coping work without claiming a fixed meaning for any figure.

Example 4 — Family/Support Theme (No Identifying Detail)

A composite, de-identified family/support scene. Roles, support, and conflict are documented descriptively and kept PHI-light — no names, ages, family configuration, or specific incidents.

D (Data): Family session used the sand tray to externalize relationship roles and supports. Participants jointly built a scene placing figures at varying distances, with one figure described by the group as "in the middle of everything" and two figures the group placed close together as "the ones who get along." Brief disagreement during placement was observed and resolved when the group renegotiated where one figure stood. Therapist summarized observed roles and reflected the group's own descriptions.

A (Assessment): Scene surfaced themes of differential closeness and a mediating role, consistent with the plan focus on communication and support. Meanings anchored to participants' stated descriptions.

P (Plan): Continue communication-focused work; revisit the renegotiation moment as a model for resolving placement disagreements at home.

What NOT to Document (The Over-Interpretation Trap)

Compare the defensible notes above with a riskier one:

Client built a wall, indicating attachment avoidance and unresolved trauma.

That sentence fails on three counts. It assigns a fixed meaning the client never gave; it reads as diagnosis-by-symbol; and it is weak under audit because it documents the therapist's inference, not the client's process or the intervention. Play therapy literature is explicit that interpretation should be delivered in a calculated way so as not to overwhelm the client, with the client's reaction used to strengthen or reject it rather than treating any reading as fixed — and that the therapist must respect and affirm cultural diversity so as to avoid erroneous assumptions across cultures (Kool & Lawver, "Play Therapy: Considerations and Applications for the Practitioner," Psychiatry (Edgmont), 2010). The goal is to hold space for and reflect the client's process, not to "decode" every action. "The dragon means anger" is banned framing.

Two more flags: do not enter PHI into a free public tool, including in a scene's labels, and never embed client identifiers in an exported file.

Documenting the Scene Image (Consent, Minimum Necessary, Storage)

If you export or screenshot the tray, treat the image as part of the record:

• Consent — get the client's (or caregiver's) consent to capture and store the scene image, consistent with your policies.
• Minimum necessary — keep only what is clinically useful; you usually do not need every iteration of the build.
• Storage — store the image in the EHR/chart, not as a download on an unmanaged device, and never email it unsecured.
• No identifiers in the file — do not put client names in the file name or in scene labels.

The free public virtual sand tray is a clinical aid, not a place to store PHI — keep names and identifying details out of it entirely.

How CoralEHR Helps

CoralEHR is an AI-first, HIPAA-compliant EHR built for private-pay behavioral-health therapists, and it signs BAAs. For play and expressive work, it keeps the scene, the note, and the goal in one record instead of scattered across screenshots and files.

Here is how the AI fits, stated plainly: CoralEHR drafts treatment-plan and note suggestions that you review, edit, and sign. There is no auto-sign and no auto-accept — an AI draft persists as a preliminary document and nothing enters the record until a licensed clinician signs it. The AI never diagnoses or decides; its guardrails forbid diagnosis, treatment, medication, and prognosis claims on every surface. No session recording is required — drafts come from the scratchpad notes you type and your structured chart fields, not a live recording. Your data trains nothing — CoralEHR runs on Anthropic's first-party Claude API, inputs are not used to train models, and logs carry hashed IDs rather than session text. The product win for sand tray work is continuity: attach the scene image to the chart, keep the note tied to a plan goal, and track themes over time in one place — instead of screenshots scattered across files.

Try CoralEHR free for 30 days, no credit card. Start a free trial, see pricing, review our BAA, or see how it fits play therapists. Want to run a session first? Use the free virtual sand tray, and reach for the Feelings Volcano or Worry Muncher for younger clients. To see it on your own caseload, book a demo.

Takeaways

• Describe the build, not the symbol — engagement, miniatures, placement, change, client words, affect, and your stance.
• Tie the note to a treatment-plan goal; the golden thread is what supports medical necessity.
• Pick SOAP, DAP, or BIRP and stay consistent — the same observations map into all three.
• Keep PHI out of public tools, scene labels, and exported file names; store any scene image in the chart with consent.
• Attribute meaning to the client's own words, never to a fixed interpretation of the object.

This article is clinician and practice-management education, not legal or clinical advice; documentation requirements vary by payer, jurisdiction, and your own policies, so verify against your contracts and licensing board.