Francine Shapiro’s 1989 study: the walk that started EMDR

In 1989, psychologist Francine Shapiro published a small controlled study testing eye movement desensitization (EMD), the technique that became EMDR, on 22 adults with trauma-related distress. One session reduced their anxiety and changed how they viewed the memory, effects that held three months later. It was a genuinely promising result, and an immediately controversial one.

Most pages about this study tell it as origin myth: a walk in the park, a chance discovery, a therapy born fully formed. The actual paper is smaller, stranger, and more interesting than that. Here’s what Shapiro tested, what she found, why other researchers pushed back almost immediately, and how a 22-person study became the seed of a therapy now used worldwide.

Shapiro’s 1989 study at a glance

• Shapiro published two papers in 1989: a full report in the Journal of Traumatic Stress, and a shorter companion piece in the Journal of Behavior Therapy and Experimental Psychiatry introducing the same new procedure.
• The main study treated 22 adults, each carrying distress from one specific memory: Vietnam combat, childhood sexual molestation, sexual or physical assault, or emotional abuse.
• Each person got a single session of EMD, then was reassessed at 1 month and 3 months.
• The technique was called EMD, not EMDR, in 1989. Shapiro added “reprocessing” in 1991 after noticing clients’ beliefs about themselves kept shifting, not just their anxiety.
• Two of the study’s biggest questions never fully closed: whether the sample really had PTSD, and whether the eye movements themselves did anything beyond the memory recall they came with.
• A 2023 paper and a 2026 follow-up both dispute the discovery story that usually opens this history.

Where did the idea for eye movements actually come from?

The familiar version goes like this: in 1987, Shapiro was walking through a park when she noticed some disturbing thoughts fading on their own, and realized her eyes were darting back and forth as it happened. She tried directing the movement on purpose, found it took the edge off other upsetting thoughts too, and started testing it on volunteers. That walk is the founding story most EMDR histories open with, including her own.

Two recent papers complicate it. Psychologist Gerald Rosen’s 2023 re-examination surfaced a 1985 magazine article in which Shapiro, two years before the walk, described teaching eye-movement exercises drawn from neuro-linguistic programming (NLP) through workshops she ran herself. A 2026 follow-up by Rosen and Loren Pankratz adds a photograph from that same period as further evidence. Separately, an earlier 1995 critique argued that the rapid, involuntary eye movements Shapiro described noticing in herself aren’t something people can typically perceive that way at all.

None of this undoes what the 1989 study measured. It does mean the tidy discovery story deserves the same scrutiny as any other claim on this page, and that “where the idea came from” and “does the technique do anything” are two separate questions worth keeping apart. For more on how Shapiro got here, see who created EMDR? The Francine Shapiro story.

What did Shapiro actually test in 1989?

The main study wasn’t a case report. It tracked three things for each of the 22 participants: anxiety tied to the memory, rated on a version of the Subjective Units of Disturbance (SUD) scale already standard in behavior therapy; how true a new, positive belief about themselves felt; and whether complaints like flashbacks, intrusive thoughts, and sleep trouble had eased. Shapiro checked all three before treatment, then again at 1 month and 3 months.

Participants received one session of EMD, holding the traumatic memory in mind while tracking Shapiro’s moving finger back and forth. That was compared against a control condition: holding the same memory in mind, without the eye movements, for a similar stretch of time.

What did the study actually find?

By Shapiro’s own report, the results were strong for something this small. Anxiety linked to the memory dropped, and stayed down at the 3-month check. Participants rated a positive belief about themselves and the incident as noticeably more true than before. And several people reported that flashbacks, intrusive thoughts, or sleep disturbances eased or stopped, changes tied to the original complaint that brought them in.

The headline claim, that one session did this, is exactly what made the paper so widely noticed and so quickly doubted. Psychotherapy results that hold up after a single hour are rare enough that they invite a second look, not automatic acceptance.

Why did other researchers question the results?

Three issues came up quickly, and a 2002 chronology of the technique’s development lays them out plainly. First, the sample was small: 22 people is enough to suggest an effect worth chasing, not enough to prove one. Second, none of the participants had PTSD formally confirmed through a structured diagnostic interview, so it isn’t fully clear how many actually had the condition EMDR later became best known for treating. Third, the control condition, sitting with a memory and doing nothing else, may simply have felt less credible or less hopeful to participants than a novel technique with a name and a rationale, a gap that alone can inflate an early result.

None of these problems were unique to Shapiro. Small, uncontrolled-for-expectancy pilot studies are how most psychotherapies start. The honest version of this story is that a promising first study is a beginning, not a verdict, on any technique, this one included.

Did the eye movements themselves matter? What later studies found

This became the field’s longest-running argument, and it started almost immediately. A 1994 dismantling study by Renfrey and Spates put 23 people with PTSD into one of three conditions: the standard EMD procedure, a version that produced the same eye movements by having people track a moving light, or a version using fixed visual attention with no eye movement at all. All three groups improved, and held those gains at follow-up. None beat the others.

A broader 1999 critical review surveyed this whole run of early-1990s dismantling studies and reached a blunt conclusion. EMDR outperformed no treatment and some weaker comparison therapies. But there wasn’t convincing evidence that the eye movements specifically were doing the work, versus the memory recall and exposure they came bundled with. That question outlived the decade: our look at working memory theory covers where the mechanism debate stands today, and is EMDR pseudoscience? covers the sharper end of that criticism.

How did one small study become a worldwide therapy?

Slowly, and only after the evidence base grew far past what Shapiro alone could produce. She renamed EMD to EMDR in 1991, adding “reprocessing” once case reports kept showing shifts in self-belief that plain desensitization didn’t explain. Independent replication picked up through the 1990s, uneven at first, and it took until 2001 for the first meta-analysis to statistically pool EMDR’s controlled trials, over a decade after that original 22-person paper.

From there the evidence compounded. By 2013, the World Health Organization named EMDR alongside trauma-focused CBT as one of only two psychotherapies it recommends for PTSD, a recognition the APA, the VA/DoD, and NICE have each since echoed in their own guidelines. Our EMDR meta-analyses roundup and is EMDR evidence-based? walk through that fuller, much larger body of research.

The honest summary: Shapiro’s 1989 study didn’t prove EMDR works. It was the small, flawed, real first data point that made asking the bigger question worthwhile.

What does this mean if you want to try bilateral stimulation yourself?

Worth saying plainly first: every person in the 1989 study was working one-on-one with Shapiro on a specific, often severe traumatic memory, in what amounted to a clinical setting. Nothing about that study, or the decades of research it kicked off, tested a self-guided app or solo practice. If what’s on your mind is trauma at that level, the research summarized here argues for a trained EMDR clinician, not for reprocessing it alone.

For everyday stress, a few things carry over from how careful researchers approach this work. Stabilize first: build a simple grounding or calm-place practice before bringing anything uncomfortable to mind, so you’re starting from steady ground. Go slow: pick one small, recent, low-intensity target instead of your heaviest memory, and keep sessions short. Know your stop-conditions: if distress climbs past a 7 out of 10 and won’t settle, stop and ground instead of pushing through; consider working with a professional if that keeps happening.

EmEase, a self-guided EMDR app, is built around that same bilateral stimulation ingredient, for practicing on your own time with everyday stress, not as a stand-in for the clinician-led work Shapiro’s 1989 subjects received. It’s a wellness practice inspired by EMDR therapy, not a test of the clinical claims this article covers.