Does EMDR work for PTSD? The trial evidence says yes — about as well as trauma-focused CBT

Eye movement desensitisation and reprocessing (EMDR) works for post-traumatic stress disorder: in the largest network meta-analysis of psychological treatments, it was among the most effective approaches at reducing symptoms, alongside trauma-focused cognitive behavioural therapy [s1]. The caveats are that the underlying evidence is graded moderate-to-low quality, and that the eye movements the therapy is named for may not be the part that helps [s1].

That first sentence is the one a reader wants, and it is less equivocal than the marketing around any single trauma therapy usually manages — because the honest version does not need to claim one therapy beat the rest.

What the largest synthesis found

The network meta-analysis pooled 90 randomised trials covering 6,560 adults and 22 interventions [s1]. Measured against a waitlist control at the end of treatment, the largest symptom reductions came from EMDR (standardised mean difference −2.07; 95% credible interval −2.70 to −1.44), combined somatic and cognitive therapies (SMD −1.69; 95% CrI −2.66 to −0.73), trauma-focused CBT (SMD −1.46; 95% CrI −1.87 to −1.05) and guided self-help (SMD −1.46; 95% CrI −2.33 to −0.59) [s1]. Non-trauma-focused CBT, CBT combined with an SSRI, SSRIs alone, unguided self-help and counselling followed [s1].

Two findings matter more than the ranking. First, EMDR and trauma-focused CBT were the two that sustained their effect at one-to-four-month follow-up, rather than fading once treatment stopped [s1]. Second, EMDR, trauma-focused CBT, guided self-help and counselling all improved the proportion of people who reached remission post-treatment [s1]. The authors' conclusion is that EMDR and trauma-focused CBT appear most effective both at cutting symptoms and at improving remission, and at holding those gains beyond the end of treatment [s1].

A second, independent review points the same way

A Cochrane review of psychological therapies for chronic PTSD reached a compatible verdict from a different set of trials: 70 studies and 4,761 participants [s2]. Individual trauma-focused CBT and EMDR were both more effective than waitlist or usual care — trauma-focused CBT with an SMD of −1.62 (95% CI −2.03 to −1.21, across 28 studies and 1,256 people), EMDR with an SMD of −1.17 (95% CI −2.04 to −0.30, across 6 studies and 183 people) [s2].

Crucially, when the Cochrane authors compared the active therapies head-to-head, they found no statistically significant difference between individual trauma-focused CBT, EMDR and stress management immediately after treatment [s2]. The consistent signal across both reviews is that the trauma-focused therapies are the front rank, and that within that rank the trials cannot reliably separate them.

What the evidence does not establish

Both teams grade their own evidence soberly. The network meta-analysis describes the quality as moderate-to-low [s1]. The EMDR estimate in the Cochrane review rests on six studies and 183 people — a thin base for a confident number, and one reason the EMDR credible and confidence intervals are wide in both reviews [s1][s2].

The reviews also do not settle the question EMDR's own theory raises: whether the lateral eye movements do anything. Neither synthesis was designed to isolate that component, and the comparative data it would take — EMDR with the eye movements against an otherwise identical protocol without them — is not what these trials tested [s1][s2]. A reader can take from this evidence that the packaged therapy works without taking from it that the mechanism its developers proposed is the reason.

Both reviews flag dropout and the question of whether findings hold in more complex presentations. The Cochrane authors note some evidence of greater drop-out in the trauma-focused arms, and the network meta-analysis authors call for more work on how severity and complexity of PTSD affect outcomes — the trials skew toward single-trauma, less complex cases [s1][s2].

Where this leaves the question

The defensible summary is narrow and useful. For adults with PTSD, EMDR and trauma-focused CBT are the psychological treatments with the strongest and most durable evidence, they perform about as well as each other where trials can compare them, and the confidence attached to all of this is moderate rather than high [s1][s2]. The case for EMDR does not require believing the eye movements are special; it rests on the outcome data, which are real.

That sits alongside a separate strand of this site's coverage — that psychedelic-assisted approaches to PTSD remain earlier in their evidence, and that for depression the main psychotherapies tend to work about equally well. This article describes evidence and is not medical advice or a treatment recommendation.