By the time a parent asks me this question, they have usually done a fair amount of reading, and they have arrived at two names. Trauma-focused cognitive behavioral therapy, which turns up everywhere and is often described as the gold standard. And EMDR, which sounds stranger and which someone has told them works quickly.
What they want is for me to say which one is better. What I can offer instead is less satisfying and considerably more useful, which is that the research separates these two treatments much less than the internet does, and that the real basis for choosing between them is not which one performs better on average. It is which one your particular child will actually do.
What each treatment asks of a child
The difference is easiest to see in terms of demand rather than theory.
Trauma-focused CBT is a structured, skills-based treatment, typically delivered across somewhere between eight and sixteen sessions, with parallel caregiver sessions running alongside the child’s. Its central component is a trauma narrative: the child builds an account of what happened, in stages, and works through it with the therapist. There is psychoeducation, work on managing distress, gradual exposure to reminders, and cognitive work on the beliefs the event produced. There is usually homework.
EMDR is an eight-phase protocol built on a different premise, which is that the memory itself is stored in an unintegrated form and can be settled through repeated brief attention to it paired with bilateral stimulation. As the World Health Organization’s guidance puts it, EMDR does not involve detailed descriptions of the event, direct challenging of beliefs, extended exposure, or homework.
Read those two paragraphs again with a specific child in mind and the choice usually starts to make itself.
What the evidence actually shows
I will give the numbers, because the summary versions on most practice websites are shaped by which treatment the practice offers.
The most rigorous recent synthesis is a 2025 network meta-analysis in JAMA Psychiatry pooling 70 randomized trials and 5,528 young patients. Trauma-focused CBT was associated with the largest reductions in pediatric PTSD, at an effect size of 1.06. EMDR came in at 0.86. Both substantially outperformed passive control. The authors noted that EMDR had insufficient long-term follow-up data to assess durability with confidence.
A 2018 meta-analysis by Lewey and colleagues in the Journal of Child and Adolescent Trauma, drawing on 30 studies, found both treatments effective at reducing post-traumatic stress symptoms, with trauma-focused CBT holding a marginal advantage.
Where the two have been compared head-to-head, the gap is smaller still. In the de Roos multicenter trial of 103 children with single-incident PTSD, EMDR and a cognitive behavioral writing therapy produced nearly identical results: effect sizes of 1.27 and 1.24 against wait-list, and remission rates of 92.5 percent and 90.2 percent. The difference was in efficiency. EMDR got there in an average of 4.1 sessions and 140 minutes of therapist time, against 5.4 sessions and 227 minutes.
On guidelines: the International Society for Traumatic Stress Studies gives strong recommendations for both, and for only one other intervention, for children and adolescents. The UK’s NICE guideline is more directive, recommending trauma-focused CBT first and advising that EMDR be considered for young people aged seven to seventeen who have not responded to or engaged with it.
The fair summary is that trauma-focused CBT has the larger evidence base and a small edge in pooled analyses, that in direct comparison the two perform very similarly, and that EMDR often arrives at the same place in less time.
So how do you actually choose
Given that the average difference is small, the sensible basis for a decision is fit. These are the considerations I weigh.
Trauma-focused CBT tends to be the better bet when the child is reasonably verbal and not actively opposed to discussing what happened, when a caregiver is available and willing to take a structured role, when homework will realistically get done, and when the child would benefit from an explicit set of skills for managing distress that will outlast the treatment. It is also well suited to situations where behavior and safety work need to run alongside the trauma work.
EMDR tends to be the better bet when the child cannot or will not produce a narrative, when shame is central to what happened, when a previous talking treatment stalled or the child refused to return, when homework is not going to happen in this household, when the target is a single discrete event, and when the child is young enough that sustained verbal work is a poor fit for their development.
Neither is the right starting point when the child is not yet safe. If the harm is ongoing, if the home is unstable, or if there is no reliably regulating adult in the picture, the first work is stabilization, and reprocessing waits regardless of which method you would eventually use.
What the first month feels like in each
Parents find the abstract comparison hard to hold, so it sometimes helps to describe how the two actually feel from the outside over the first several weeks.
In trauma-focused CBT the early sessions are largely educational and skills-based, and to a parent watching from the waiting room it can look reassuringly like ordinary therapy. The child learns about how trauma affects the body, practices ways of managing distress, and does exercises at home. The narrative work comes later and is built in pieces. Because the difficult part arrives some weeks in, families sometimes get an early sense that things are going well and then hit a harder stretch around the middle of treatment.
In EMDR the shape is different. The preparation phase is doing real work but can look to a parent like very little is happening, which is the most common reason families lose faith in the first fortnight. Then processing begins and the change can be abrupt. Children frequently come out of a session having visibly shifted, and it is common for a child to be tired or emotionally raw that evening and noticeably better within a day or two.
Neither pattern is better. But knowing which one you are in prevents a family from concluding that a treatment is failing when it is simply at the part that does not look like much.
What matters more than the choice
Here is the thing practice websites tend not to say, including, until this sentence, mine.
The differences between these two treatments are smaller than the differences between clinicians. Whether the therapist has trained specifically in child and adolescent work rather than learning the protocol in its adult form. Whether they can build enough of a relationship with your child that your child comes back. Whether they know how long to spend in preparation before targeting anything, and can tolerate spending it. Whether they involve you usefully rather than either excluding you or making you a co-therapist.
A well-trained clinician using the treatment that fits your child will outperform a poorly matched one using the treatment with the marginally better meta-analysis. If you are choosing between two clinicians who offer different modalities, choose on the clinician.
Doing both is not a failure
Parents sometimes treat these as a fork in the road where picking wrong means losing months. In practice, they combine perfectly well.
A child who stalls in trauma-focused CBT at the narrative stage can move to EMDR for the memory itself and return to the skills work afterward. A child who processes a single event quickly with EMDR may still benefit from the caregiver work and the coping skills that trauma-focused CBT does explicitly. Sequencing them is common and is not evidence that the first attempt failed.
Our EMDR clinician at Balanced Mind of New York works with children and adolescents from age five upward, on the Upper East Side and virtually across New York State. If you are weighing the options for your child, you are welcome to reach out for a consultation, including one where we conclude that a different approach is the better fit.
Frequently asked questions
Which one works faster?
EMDR generally requires less therapist time for the same result. In the one large head-to-head trial in children, EMDR averaged 4.1 sessions and about 140 minutes against 5.4 sessions and 227 minutes for the comparison treatment, with nearly identical outcomes. For complex or repeated trauma, the difference matters much less, because most of the work in either treatment sits in preparation.
Is trauma-focused CBT better than EMDR?
It has more trials behind it and a small edge in pooled analyses. In direct comparison, the two perform similarly. Describing either as clearly superior overstates what the research supports.
Why does NICE recommend trauma-focused CBT first?
Guideline bodies weigh volume and consistency of evidence, and trauma-focused CBT has been studied more extensively in children. NICE advises considering EMDR for those aged seven to seventeen who have not responded to or engaged with it. That ordering reflects the state of the research rather than a judgment that EMDR is ineffective.
My child refuses to talk about what happened. Does that rule out trauma-focused CBT?
Not automatically, since the narrative work is built gradually and a skilled clinician does not start there. But sustained refusal is one of the clearer indications for EMDR, which does not require a detailed account of the event.
Does my child have to relive the trauma in either one?
Neither treatment works by making a child endure the memory unprotected, though they approach it differently. Trauma-focused CBT builds toward the trauma narrative gradually, with skills taught first specifically so the child can tolerate it. EMDR involves brief, repeated contact with the memory rather than sustained exposure, and does not require the child to describe it. Both should feel difficult at points, and neither should feel like being thrown back into it.
Can we switch if the first one is not working?
Yes, and it is common. What is worth checking before switching is whether the treatment has actually had a fair run, since the preparation phase in both approaches can look like nothing is happening when the necessary groundwork is being laid.