EMDR Treatment for Post-Traumatic Stress: What Trials Show

Eye movement desensitization and reprocessing (EMDR) is a post-traumatic stress treatment in which a person holds a distressing memory in mind while following a back-and-forth movement or sound, about 30 seconds at a time^[7]. The 2023 Department of Veterans Affairs and Department of Defense (VA/DoD) guideline names it, with cognitive processing therapy and prolonged exposure, as one of three individual psychotherapies it recommends most strongly for PTSD^[1].

This article covers how EMDR works, how it compares with cognitive behavioral therapy and where its evidence is weak. For how many people live with the condition, see the PTSD statistics page; for an overview of every option, see PTSD treatment options.

Psychologist Francine Shapiro introduced the method in a 1989 study of 22 people with trauma-related symptoms, reporting that a single session reduced the distress tied to their memories through a 3-month follow-up^[8]. Trials from the 1990s were less rigorous and produced conflicting results, while more recent trials using a 3-month protocol support its efficacy^[9].

EMDR's working model holds that PTSD symptoms come from trauma memories that were never adequately processed, and that bilateral stimulation helps integrate those memories into the wider memory system^[9]. That framing overlaps with research on how trauma affects the brain, though the mechanism itself is disputed.

EMDR differs from other trauma-focused therapies mainly in what it leaves out. It does not include extended exposure to the memory, detailed descriptions of the trauma, challenging of beliefs or homework assignments^[4].

Treatment starts with history-taking, an explanation of the method and techniques for managing stress reactions during sessions^[9]. The therapist and patient also discuss how ready the patient is to focus on trauma memories before any processing begins^[7].

In the assessment step, the patient picks a target image from the memory, often the most distressing one, plus a negative belief about the self such as "I am powerless," and rates their current distress^[9]. They then name a belief they would prefer to hold, such as "I now have choices," and rate how true it feels while recalling the memory^[9].

During desensitization, the therapist leads sets of bilateral stimulation while the patient attends to the memory and notices whatever thoughts, emotions and body sensations surface^[9]. Sets continue until distress is greatly reduced and the preferred belief feels true, and a body scan then checks for leftover physical tension^[9].

The same steps are repeated for current situations that trigger distress and for imagined future ones^[9]. In most cases the patient is not asked to describe the trauma in detail, though they are asked to think about it during sessions^[7].

How many sessions does EMDR take?

The VA's National Center for PTSD describes a typical course as about 3 months of weekly sessions lasting 50 to 90 minutes, and says many people notice improvement after a few sessions^[7]. The American Psychological Association (APA) describes EMDR as usually delivered once or twice a week for a total of 6 to 12 sessions^[4].

EMDR can also be delivered by video, using apps or websites that guide eye movements or play tones^[9]. The VA/DoD guideline strongly recommends video delivery of its recommended therapies when a therapy has been validated for video or other options are unavailable^[1].

The broadest comparison is a 2020 network meta-analysis of 90 trials, 6,560 adults and 22 interventions, which found EMDR and trauma-focused cognitive behavioral therapy (TF-CBT) most effective at reducing symptoms and improving remission^[2]. EMDR's pooled effect against a waitlist was a standardized mean difference of −2.07, compared with −1.46 for TF-CBT, though the authors rated the evidence moderate to low quality^[2].

A 2025 review limited to trials published since 2018 identified 16 randomized controlled trials (RCTs) with 1,031 adults^[3]. EMDR outperformed waitlist or usual care after treatment (standardized mean difference 1.86) and at a 3-month follow-up (2.44), and it did not differ significantly from TF-CBT^[3].

The same review flagged extremely high heterogeneity, meaning results varied so much that one trial's effect could be at least 50 times another's^[3]. Most of its trials ran outside Europe, and the authors called for further studies before definitive policy recommendations^[3].

The VA's PTSD Trials Standard Data Repository lists 38 RCTs of EMDR^[9]. In a head-to-head trial of 88 people given 8 weeks of treatment, EMDR produced more lasting reductions in PTSD and depression symptoms than the antidepressant fluoxetine, mainly among people whose trauma began in adulthood^[10].

Cognitive behavioral therapy (CBT) for PTSD is a family of treatments rather than a single method. The APA guideline's first-line options are all CBT variants: general CBT, cognitive processing therapy (CPT), which challenges unhelpful beliefs about the trauma over about 12 sessions, and prolonged exposure (PE), which has patients gradually approach avoided memories and situations^[4].

The two major U.S. guidelines place EMDR differently. VA/DoD puts it beside CPT and PE in its strongest category, while the APA guideline, updated in 2025, gives it a conditional, second-line rating^[1][4].

APA says a conditional rating means the evidence may be less strong, the balance of benefits and harms less favorable or the treatment less applicable across settings and groups^[4]. It does not mean the therapy was found ineffective.

Cognitive therapy for PTSD, which VA/DoD calls Ehlers cognitive therapy, works on the pessimistic evaluations and memories a person attaches to the trauma^[4]. VA/DoD gives it a weaker "suggest" rating, alongside present-centered therapy and written exposure therapy, and notes as a limitation that all of its studies were run by the treatment's developers^[1].

Head-to-head comparisons mostly end in a tie. An analysis of raw patient data from 8 trials and 346 patients found no significant difference between EMDR and other psychological therapies in symptom reduction, response, remission or dropout^[11]. The 2025 review reached the same conclusion for EMDR versus TF-CBT and noted that EMDR courses were often shorter^[3].

VA/DoD recommends individual trauma-focused psychotherapy over medication and, when medication is used, recommends paroxetine, sertraline or venlafaxine^[1]. Only sertraline and paroxetine are FDA-approved for PTSD^[4].

The dispute is less about whether EMDR helps than about why. The VA's own provider summary notes continuing controversy over its mechanism of action and whether bilateral stimulation is needed at all^[9].

A 2001 meta-analysis of 34 studies found that EMDR beat no treatment but was no more effective than other exposure techniques, and that adding eye movements made no measurable difference^[5]. Its authors concluded that the eye movements "integral to the treatment, and to its name, are unnecessary"^[5].

A 2013 meta-analysis reached a different answer^[6]. Across 15 clinical trials comparing EMDR with and without eye movements, the eye movements added a moderate effect (Cohen's d = 0.41), and across 11 laboratory studies the added effect was large (d = 0.74)^[6].

The VA cautions that the 2013 analysis included studies beyond PTSD that relied mainly on self-report measures^[9]. The evidence base has other weak spots: trials from the 1990s produced conflicting results, and the 2025 review found extremely high variation between trials^[9][3].

Discomfort is the expected side effect. The VA tells patients they may feel uncomfortable when focusing on trauma memories or beliefs, that these feelings are usually brief and that most people who complete EMDR find the benefits outweigh the initial discomfort^[7].

Serious harms appear uncommon in trials. VA/DoD judged the potential harms of CPT, EMDR and PE, such as adverse events, to be small, and the 2025 review reported a very low discontinuation rate and rare adverse events for EMDR^[1][3].

EMDR does not work for everyone. In the fluoxetine comparison trial, 75.0% of patients whose trauma began in adulthood reached asymptomatic functioning 6 months after EMDR, compared with 33.3% of those whose trauma began in childhood, and no one in the fluoxetine group reached that level^[10].

In the pooled patient-data analysis, men were more likely than women to drop out of EMDR, and unemployed participants had higher symptom scores after treatment^[11].

A second diagnosis is not automatically a barrier. In a trial of 155 patients with both PTSD and a psychotic disorder, EMDR and prolonged exposure each outperformed a waitlist, with results maintained at 6 months^[9]. VA/DoD found that co-occurring conditions, including substance use disorders, did not appear to change the safety or effectiveness of recommended PTSD therapies^[1].

EMDR is offered in many VA medical center PTSD programs, and smaller VA facilities can deliver it by videoconference from a clinician at another site^[7]. The research base for this group is thin, however, since most EMDR studies have been done in civilians^[9].

A review found 4 RCTs of EMDR in military or veteran samples, all published before 2000, and only one delivered a full course of treatment, while the rest lasted 1 to 3 sessions^[9]. Those trials found EMDR improved symptoms about as much as the comparison conditions, which included exposure therapy, usual care and biofeedback^[9].

For data on how PTSD affects service members and veterans, see PTSD among veterans and veterans and PTSD.