Read time: 9 minutes
You can name the 8 phases of EMDR therapy. You learned them in Basic Training, you have the manual on a shelf somewhere, and if a supervisee asked, you could recite the list. But there is a difference between naming the phases and knowing, in the middle of a hard session, exactly what job each phase is doing and why the order matters.
That second kind of knowing is what this guide is for.
This is a working review for trained clinicians, written the way we wish someone had laid it out after Basic Training: each phase, the clinical task it exists to accomplish, and the place where that task most often slips. If you work with clients and not just concepts, read it with your current caseload in mind.
(If you landed here as a client or a therapist curious about what EMDR is, you are welcome too. You will leave with a clear picture of how the therapy is structured and why it works the way it does.)
It is tempting to treat the eight phases as scaffolding: useful while you learned, safe to loosen once you found your own style. The research says otherwise. Hase (2021) argues that the structure of EMDR therapy, including the standard protocol and its procedural steps, is a core element of the approach itself, and that following that structure is tied to treatment effect. The most recent state of the science review (de Jongh et al., 2024) reaches a similar place from the evidence side: the trials behind EMDR’s standing as a first-line PTSD treatment were run on the protocol as designed.
The phases also carry the theory. EMDR rests on the Adaptive Information Processing model, the idea that present symptoms are driven by maladaptively stored memory networks, and that the brain can integrate those memories when processing is activated under the right conditions. Every phase maps onto that model. Phases 1 and 2 locate the networks and build the capacity to approach them. Phase 3 activates a specific network. Phases 4 through 6 process it to adaptive resolution. Phases 7 and 8 close the container and check the work.
Seen that way, the phases stop being a compliance exercise. They are the AIP model translated into clinical behavior, in order. When a session stalls, the most useful diagnostic question is usually some version of: which phase am I actually in right now, and did I finish the one before it?
The job: build a case conceptualization through the AIP lens and select targets worth processing.
Phase 1 is where EMDR treatment planning lives, and it is more than intake. You are listening for the memory networks underneath the presenting problem: the touchstone events, the present triggers that activate them, and the future situations where the client wants to respond differently. A good Phase 1 produces a target sequence, which means that by the time you reach Phase 3 you are choosing from a plan rather than improvising in the room.
Where it slips: clinicians collect a thorough history and never convert it into targets. The result is EMDR that starts wherever last week’s crisis points, which is one of the most common reasons processing later loops or stalls.
The job: build enough stabilization, skills, and shared understanding that the client can process with dual awareness.
Preparation covers psychoeducation about EMDR and the AIP model, informed consent, resourcing (calm place, container, and whatever this particular nervous system needs), and testing bilateral stimulation. It is also where the therapeutic relationship earns its keep. Hase and Brisch (2022) make the case that the relationship is itself a core element of EMDR therapy, and Phase 2 is where most of that foundation gets poured.
Where it slips: in both directions. Some clinicians rush preparation and meet dissociation or flooding in Phase 4. Others camp in Phase 2 for months with clients who were ready long ago, usually because the therapist, and sometimes the client, is anxious about reprocessing. Readiness criteria exist so this call does not rest on anxiety.
The job: activate one specific memory network and take baseline measurements before processing begins.
Phase 3 is a structured sequence: image, negative cognition, positive cognition, VoC rating, emotions, SUD rating, body location. Each component has a purpose. The image and negative cognition light up the network. The positive cognition points toward adaptive resolution. The VoC and SUD give you numbers to compare in Phases 4 and 5. The order is deliberate, moving the client from narrative into activation.
Where it slips: this is the phase clinicians most often report fumbling, and the one where scripts get quietly abbreviated or reordered. Skipping the VoC, accepting a vague negative cognition, or taking the SUD before the emotion step all weaken your ability to track whether processing is working.
The job: reprocess the target memory with sets of bilateral stimulation until distress resolves, ideally to a SUD of 0.
This is the phase most people picture when they picture EMDR. Your work here is mostly disciplined restraint: brief check-ins between sets, staying out of the way of spontaneous processing, and intervening only when processing is blocked. When it is blocked, you have decisions to make. Is this looping that needs an interweave? Is the client outside their window of tolerance? Do you need to return to a skipped piece of Phases 1 through 3? Knowing your interweave options, and when constricted approaches like EMD or EMDr serve the client better than the full protocol, is what separates confident Phase 4 work from white-knuckling.
Where it slips: talking too much between sets, abandoning targets before resolution, and pushing through dysregulation instead of adjusting. All three are drift wearing a clinical costume.
The job: strengthen the positive cognition until it feels fully true when paired with the target memory.
You return to the positive cognition from Phase 3, confirm it still fits (processing often produces a better one), and run sets of bilateral stimulation until the VoC reaches 7, or as close as is ecologically valid. Installation is quick when Phase 4 was complete. When it drags, that is information: a VoC that will not budge usually means residual material or a blocking belief, and sometimes a feeder memory your Phase 1 map did not catch.
Where it slips: treating installation as a formality and accepting a VoC of 5 because the hour is ending. The adaptive network you are consolidating here is what the client carries out the door.
The job: confirm that resolution has reached the body.
The client holds the target and positive cognition in mind and scans for residual sensation. Anything that surfaces gets processed with further sets. This phase takes minutes, and it exists because trauma that resolves cognitively can remain physiologically active. A clear body scan is your evidence that the network has actually integrated.
Where it slips: it gets skipped. Of all eight phases, this is the one clinicians most often admit to dropping when time runs short, which quietly removes the protocol’s built-in verification step.
The job: return the client to equilibrium before they leave, whether or not the target is complete.
Closure covers grounding, containment for incomplete sessions, briefing the client on what processing between sessions can look like, and a plan for managing it. Done well, Phase 7 is why clients come back willing to go again.
Where it slips: running processing to the final minute and compressing closure into a goodbye. Clients who leave activated and spend three days flooded are the clients who start canceling.
The job: open the next session by checking what held.
You reassess the previous target (SUD, VoC, body), ask what surfaced during the week, and let the answers steer the treatment plan: resume the target, move to the next one, or follow new material. Phase 8 is the loop that turns individual sessions into a course of treatment.
Where it slips: skipping straight to a new target because last week ended clean. Reevaluation is also where the three-pronged protocol is tracked, which is the next thing worth reviewing.
Full EMDR treatment addresses all three prongs for each presenting issue. First the past memories that set the network down, then the present triggers that still activate it, then a future template in which the client rehearses responding from the adaptive network. Most incomplete EMDR is incomplete here. The past prong resolves, symptoms improve, and treatment ends before present triggers and future templates are processed, leaving the client with relief but without the full generalization the protocol is designed to produce. If your treatment plans tend to trail off after the past prong, this is one of the highest-yield things to tighten.
Clinicians across modalities gradually stop delivering evidence-based treatments as designed, even when they have the training and the tools. Waller and Turner (2016) describe how this happens through ordinary clinical pressures: therapist anxiety, client avoidance, time constraints, and the slow substitution of talking about distress for techniques that reduce it. None of it requires a bad clinician. It requires a busy one.
In EMDR, drift has recognizable shapes, and you have probably seen several in this article already: history without a target sequence, preparation without end, an abbreviated Phase 3, chatty Phase 4s, skipped body scans, vanishing future templates. Each one feels minor on its own. Together they produce the sessions that leave you wondering why EMDR used to work better than it does now.
The repair is unglamorous and effective: review the model, re-anchor each phase to its job, and practice the pieces that went soft. We wrote more about the pattern itself in Therapist Drift in EMDR: What It Is and How to Reverse It, and about when a structured review makes sense in EMDR Refresher: How to Get Back to Clean, Confident Protocol.
Live training · January 21 & 22, 2027 · 10:00 AM–2:30 PM Central · 8 CEs
If this article named a few phases that have gone soft in your practice, you are in good company, and the fix is simpler than retraining. Over two live sessions, Crystal Wildes, PsyD, LP walks you back through the AIP model, EMDR treatment planning, all 8 phases, and what to do when processing blocks or loops. You leave with the protocol sharp again and tools you can use in your next session. Lifetime on-demand access is included, so the review stays on your shelf for the cases that need it.
8 CEs · EMDRIA, ACE, NBCC, APA · Live online + lifetime on-demand access
I’m Getting Back to Clean EMDR →Yes, within each target. The sequence is designed so that each phase creates the conditions the next one requires. Across a course of treatment the work is more fluid: Phase 8 can send you back to history taking or preparation whenever the clinical picture calls for it.
It varies by client. Phases 1 and 2 may take one session or many, especially with complex trauma. Phases 3 through 7 often occur within a single session per target, and Phase 8 opens every subsequent session. The protocol sets the order of the phases, and the client’s nervous system sets the pace.
Phase 3 is the most commonly fumbled, since the assessment sequence is precise and easy to abbreviate under pressure. Phase 4 generates the most clinical anxiety, because blocked processing, looping, and dysregulation all demand real-time decisions that Basic Training can only partially prepare you for.
The 8 phases describe how each individual target is processed from history to reevaluation. The three-pronged protocol describes which targets a full course of treatment must cover: past memories, present triggers, and future templates. Complete EMDR treatment requires both working together.
No. If you completed an EMDRIA-approved Basic Training, what you need is a structured review, since skills fade with low caseload volume or accumulated drift rather than disappearing. A refresher course rebuilds precision on the full 8-phase model without repeating foundational training you already hold.
de Jongh, A., de Roos, C., & El-Leithy, S. (2024). State of the science: Eye movement desensitization and reprocessing (EMDR) therapy. Journal of Traumatic Stress, 37(2), 205–216. https://doi.org/10.1002/jts.23012
Hase, M. (2021). The structure of EMDR therapy: A guide for the therapist. Frontiers in Psychology, 12, Article 660753. https://doi.org/10.3389/fpsyg.2021.660753
Hase, M., & Brisch, K. H. (2022). The therapeutic relationship in EMDR therapy. Frontiers in Psychology, 13, Article 835470. https://doi.org/10.3389/fpsyg.2022.835470
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. https://doi.org/10.1016/j.brat.2015.12.005