July 19, 2026
11 min read
By Albert Wong, PhD · Clinical Psychologist
The short answer
Document the EMDR phase by name, the target at a headline level ("index incident from 2019," not the details), the negative and positive cognitions, the SUD and VoC scores at the start and end of the session, the bilateral stimulation modality, whether the session was complete or incomplete (and what closure exercise you used), and the client's stability at departure. Bill it as standard psychotherapy — 90834 or 90837 — and tie the scores to your treatment-plan goals. Structure plus scores plus safety. The trauma narrative itself stays out.
The session was the best clinical work you did all week. Twenty minutes of sets. The moment where your client's shoulders finally dropped. Tears, then something quieter than tears. A SUD that had been anchored at 8 for months finally slipping to 3. You sat with someone while a memory that had been running their life for six years loosened its grip.
And now the note field is blank, the cursor is blinking, and nothing in your graduate documentation training covered what to call any of it. "Client moved her eyes while thinking about the accident" is technically accurate and clinically absurd. So a lot of EMDR clinicians write notes that are either three vague lines ("processed trauma material, client tolerated well") or a transcript of the session that no one should ever have written down. Both are problems. Here is the middle path.
Picture the person who will actually read your note: a utilization reviewer who may have never sat through an EMDR training, skimming for evidence that a covered psychotherapy service happened. They are reading your note from the shore. They never saw the water. If the note says "bilateral stimulation, SUD 8 to 3" and nothing else, they see jargon and numbers with no frame — nothing they recognize as psychotherapy. Notes like that get flagged, and flags become records requests, and records requests become clawbacks.
The instinct, when a note gets questioned, is to add more detail. That's the wrong fix. The reviewer doesn't need more of the session; they need the session translated. EMDR is not a mysterious ritual — it is a structured, protocol-driven, evidence-based psychotherapy with named phases, defined measures, and a clear theory of change, developed within Francine Shapiro's Adaptive Information Processing model. Your note fails when it hides that structure. It passes when it shows it.
You already know these terms. The move is realizing they are not just clinical tools — they are documentation, ready-made. Standard EMDR runs through eight phases, and naming the phase you worked in is the single fastest way to show structure:
Then the measures. The negative cognition (NC) and positive cognition (PC) are the beliefs under the memory — "I'm powerless" and "I have choices now." The Validity of Cognition (VoC) scale rates how true the positive belief feels, 1 to 7. The Subjective Units of Disturbance (SUD) scale rates distress, 0 to 10. And here is the part I want every EMDR clinician to appreciate: these scores are a gift to your documentation that almost no other modality gives you. Most psychotherapy notes have to argue that progress happened. Yours can show it in numbers, every single session. SUD and VoC are your depth soundings — objective markers a reviewer can track across a whole course of treatment without understanding a single thing about eye movements.
Round out the vocabulary with the bilateral stimulation (BLS) modality — eye movements, alternating taps, or auditory tones — a process-level word about what came up ("channels included grief-related associations and somatic activation"), and the closure state: complete or incomplete, and what you used to close.
Put together, an EMDR progress note has seven load-bearing pieces: the phase, the target at a headline level, the cognitions, the scores, the BLS modality, the closure state, and the client's stability at departure. Here's what that looks like assembled — a fictional composite of a phase 4 session:
Worked example — EMDR phase 4 progress note (fictional composite)
Dx: PTSD (F43.10). 53-minute individual psychotherapy session (90837); session 11 of the current treatment plan.
Interventions: EMDR desensitization phase (phase 4) targeting index incident from 2019. NC: "I'm not safe." PC: "It's over; I'm safe now." Pre-session SUD 8/10; VoC 2/7. Reviewed current stressors and confirmed stability and readiness to continue reprocessing. Conducted approximately 14 sets of bilateral stimulation (eye movements) per standard protocol, returning to target between channels. Client processed associated material at a process level, including somatic activation (chest tightness) and grief-related associations. SUD decreased from 8 to 3 over the session.
Closure: Session incomplete — target not fully desensitized; installation deferred. Incomplete-session closure protocol completed: container exercise and calm-place resourcing. Client reoriented to the present with affect visibly settled.
Risk/stability: No suicidal or homicidal ideation reported or observed. Client demonstrated adequate affect regulation at close and left the session calm, with between-session plan reviewed (brief log of any disturbance; use calm place) and crisis resources confirmed.
Plan: Continue phase 4 on this target next session; reevaluate SUD/VoC at the start of session per protocol. Progress consistent with treatment-plan goal of reducing trauma-related distress and avoidance.
Read that note as a reviewer. You may not know what a "container exercise" is, but you can see a diagnosis, a named evidence-based protocol, a measurable within-session change, a safety assessment, and a plan tied to treatment goals. That is a psychotherapy note. It took its author about four minutes.
Notice what the target description does: "index incident from 2019" is a headline, not a story. That's deliberate, and it's the next point.
The graphic trauma narrative does not belong in the progress note. Not the details of the assault, not the imagery that surfaced in set nine, not the verbatim things your client said through tears. The progress note is a record that can be subpoenaed, requested by the client, audited by a payer, and read by future providers. Trauma clients deserve a chart that proves treatment happened without re-recording the injury — the full argument is in our guide to trauma-informed documentation.
If you personally need somewhere to keep process detail — your hypotheses about a channel, a fragment you want to remember for case consultation — that is what psychotherapy notes (in the formal HIPAA sense) are for: a separate record, kept apart from the chart, with stronger protections and no billing role. The distinction has real rules, and we've written them up separately. The one-line version: the progress note carries structure, scores, and safety; anything richer lives in the other record or nowhere at all.
The incomplete session. Reprocessing doesn't respect the clock, and an unfinished target is normal, not a failure. When a squall blows up mid-crossing, you don't pretend you reached the far shore — you make for the nearest safe harbor and log it that way. Document exactly that: "session incomplete, target not fully desensitized," the closure protocol you used (container, calm place, grounding), and the client's state at departure. An incomplete session that is closed well and documented is unimpeachable. An incomplete session with no closure documentation reads like a client sent home mid-storm.
The history and preparation sessions. Phases 1 and 2 can run for weeks with anxious or complex-trauma clients, and clinicians sometimes under-document them because "nothing happened yet." Everything happened. History-taking and treatment planning is assessment-driven psychotherapy; resourcing is skills-based intervention. Write the rationale: "Phase 2 preparation session: taught container and calm-place skills to build affect-regulation capacity required before trauma reprocessing; client able to self-soothe to SUD 2 within session." That is a defensible, medically necessary session — you're charting the harbor before you sail, and the chart is billable work.
The long session. EMDR reprocessing regularly wants more than 45 minutes, which is why many EMDR clinicians live in 90837. There is no special EMDR code — it bills as standard psychotherapy, 90834 or 90837 by time — so the only extra documentation burden is the time itself: note start and stop times or total minutes, since 90837 requires 53 minutes or more. If you're unsure which code fits your sessions, our 90837 versus 90834 guide covers the thresholds and the payer quirks. And yes, EMDR works over telehealth — virtual BLS tools are mature now — in which case place-of-service codes and modifiers follow the same rules as any telehealth psychotherapy session.
The reevaluation note. Phase 8 is a documentation moment clinicians skip because it feels like housekeeping. Don't skip it. "Reevaluation of target from previous session: SUD 3 at last close, 2 today; VoC 5; gains holding" is two lines that prove treatment effects are durable — exactly the question a reviewer deciding on continued authorization is asking.
Every note ties back to two anchors: a diagnosis that supports trauma treatment — PTSD, other trauma- and stressor-related disorders — and a treatment plan whose goals your scores can measure. This is where EMDR documentation gets genuinely easy, because the modality hands you goal language most therapists have to invent: "Reduce disturbance on index target from SUD 8 to 2 or lower; increase VoC on adaptive cognition from 2 to 6 or higher." Write the goals in score terms in the treatment plan, report the scores in every progress note, and your chart tells one continuous, auditable story of medical necessity from intake to discharge. Risk and stability at close belong in that story every single time — trauma work destabilizes by design, and the payer, the licensing board, and your own future self all want to see that you checked the client's footing before they walked out the door.
Everything above fits in a template: phase, target headline, NC/PC, SUD and VoC fields, BLS modality, closure state, risk at close, plan. Once those prompts exist in your note form, the hardest sessions become the easiest notes — you fill in eight fields while the session is still warm instead of staring at a blank box deciding what EMDR is. Practice Harbor's note templates are shapeable to your modality, so an EMDR clinician can build exactly that form once and have every reprocessing session walk itself into a reviewer-ready note.
Build note templates around your modality — EMDR phase, SUD/VoC, closure, and risk prompts included — so documentation keeps pace with the clinical work. Free for pre-licensed clinicians, $19/mo licensed.
Show the structure of a recognized psychotherapy: name the EMDR phase (for example, phase 4 desensitization), identify the target at a headline level, record the negative and positive cognitions, report SUD (0–10) and VoC (1–7) scores as objective progress markers, note the bilateral stimulation modality, state whether the session was complete or incomplete and what closure exercise you used, and document the client's risk and stability at the end of the session. Tie the scores to treatment-plan goals for a trauma-related diagnosis.
There is no EMDR-specific CPT code. EMDR bills as standard individual psychotherapy — 90834 (38–52 minutes) or 90837 (53+ minutes). Because reprocessing sessions often run long, many EMDR clinicians bill 90837; document session start and stop times or total minutes to support the code. Telehealth EMDR uses the same codes with the applicable place-of-service code and modifier.
No. The progress note needs the target at a headline level ("index incident from 2019"), not graphic detail. Progress notes can be requested by clients, payers, courts, and future providers, so a detailed trauma narrative re-records the injury without adding clinical or billing value. If you need to keep process detail for your own work, use psychotherapy notes kept separate from the chart, which carry stronger privacy protections.
State plainly that the session was incomplete — the target was not fully desensitized — then document the closure protocol you used (container exercise, calm place, grounding), the closing SUD if taken, the client's affect regulation and safety at departure, and the plan to resume reprocessing and reevaluate SUD/VoC next session. Incomplete sessions are a normal part of EMDR; what reviewers and boards look for is evidence the client was returned to stability before leaving.