EVIDENCE UPDATE · 2026

EMDR research in 2026: what changed and what did not

A new publication does not always carry the same weight: a guideline, meta-analysis, definitive trial, feasibility study and protocol answer different questions. This update summarises recent signals without turning headlines into recommendations.

Written and reviewed by EMDR Match Editorial Team27 August 2026Identified clinical sources · no individual clinical review claimed

Short answer

Research published through August 2026 reinforces that EMDR may reduce PTSD symptoms compared with waiting list or usual care, while effects are smaller and comparable when active treatments are used. New publications broaden contexts and mechanisms without establishing universal superiority or automatic transferability. Important gaps persist: adverse effects are sparsely measured and many populations remain underrepresented.

PTSD: the comparator changes the conclusion

A review of trials against waiting list found favourable signals for symptoms and diagnostic remission, but rated certainty low or very low and noted missing outcomes on quality of life, acceptability and adverse effects.

The meta-analysis published in June 2026 pooled eleven trials: effects were larger against passive controls and smaller and comparable against active interventions, with methodological concerns in most studies. Beating a waiting list does not establish clinical superiority.

Safety: missing data is not evidence of no harm

The 2026 review examined 51 randomised trials included in recent meta-analyses. Only nine mentioned adverse effects and only one used systematic assessment.

Reported effects were generally mild and temporary, but missing criteria and monitoring prevent a reliable risk estimate. Informed consent and monitoring should not rely only on dropout or anecdotes.

New populations without universal conclusions

A trial of 96 refugee young people found PTSD reduction against waiting with both EMDR and KIDNET and a clinician-rated difference favouring EMDR, but not across every outcome. Another trial of 209 people studied EMDR, exposure and imagery rescripting alongside substance-use treatment and reported secondary psychosocial outcomes.

A preliminary Jordan trial compared nurse-delivered EMDR with waiting in 70 people and followed participants for only one month. These studies broaden context and access; they do not decide which professions may provide psychotherapy in each jurisdiction or replace active trials, replication and long follow-up.

August: mechanisms, representation and protocols

An online experiment with 560 participants found that recall plus eye movements reduced vividness and unpleasantness more than recall alone for memories dominated by fear, shame, anger or sadness. This was a memory proof-of-principle study, not a clinical EMDR trial or evidence of effectiveness for each emotion. An intensive study in 38 veterans reported improvement at three months, but lacked a randomised comparison and combined EMDR with additional supports.

A review of 519 eligible reports of randomised trials of CPT, prolonged exposure or EMDR found only ten reports that recorded sexual orientation and gender identity and enrolled at least one sexual- or gender-minoritised participant; it cannot estimate EMDR-specific effectiveness in those groups. Another paper describes a planned trial in 40 children with sickle cell disease: as a protocol, it provides no pain outcome yet. None of these signals changes a clinical guideline by itself.

FAQ

Frequently asked questions

Does 2026 research show EMDR is better than CBT?

No. A recent review found no significant difference from trauma-focused CBT, while both improved PTSD symptoms.

Are EMDR adverse effects well understood?

Not sufficiently. One review found that most analysed trials did not report them and almost none assessed them systematically.

Does one new trial make a use recommended?

No. Design, size, comparator, replication, certainty, harms and guidelines matter. August publications include proof-of-principle, preliminary studies and a protocol; none establishes a new effective use by itself.

Sources and scope

Informational content based on guidelines and professional bodies. It does not diagnose, recommend individual treatment or replace clinical assessment.