Updated October 20, 2026

OCD Research
Updates: What's
New, What It Means

By Justin La Rose, RP, M.Psy · Registered Psychotherapist (CRPO) · Clinical Director, The OCD Clinic

Evidence last reviewed: October 2026

The biggest OCD research news of 2026 so far: a BMJ review confirmed ERP and SSRIs as the first-line treatments, and the first randomized trial of repeated psilocybin doses showed promise in just 15 people. Below, each study gets the same five questions: what they did, what they found, what it doesn't show, my take as an OCD therapist, and whether it should change anything for you.

Here's the thing about OCD research headlines. A small trial is a witness, not a verdict. Worth hearing, never the whole case. I read the papers so a headline doesn't do your thinking for you, and every entry links to the study. Don't trust me. Read them. And remember what research is: an average of a lot of people. Your therapy is a study of one.

The short answer

  • First-line treatment hasn't changed: A 2026 BMJ review of fifteen years of evidence still puts ERP and SSRIs first (Abramowitz et al., 2026).
  • Psilocybin has its first randomized trial: In 15 people, repeated psilocybin doses reduced OCD symptoms where an active placebo didn't, and the authors call for larger trials (Moreno et al., 2026).
  • I-CBT is promising but unproven against ERP: In a 197-person trial, I-CBT and CBT both helped, but whether I-CBT is as good remained inconclusive (Wolf et al., 2024).
  • Postpartum OCD can be predicted in pregnancy: Stronger obsessive beliefs at 20 weeks of pregnancy predicted a higher chance of OCD six months after birth (Abramowitz et al., 2025).

Which OCD studies matter most right now?

StudyPublishedType of evidenceChanges treatment now?
Does psilocybin work for OCD? The first randomized trial (2026)March 2026Small randomized trial: an early signalNo, research only
What is the best OCD treatment right now? The BMJ review (2026)February 2026Review of the evidence: strong for averagesConfirms the current order
What do the IOCDF's 2025 OCD treatment guidelines recommend?2025 (guide read October 2026)Guideline: strongest for what to do firstConfirms the current order
Can beliefs in pregnancy predict postpartum OCD? (2025)September 2025Prospective cohort: shows prediction, not causeEarlier screening, maybe
Is I-CBT as good as ERP for OCD? (2024)October 2024Large randomized trial: good, but one studyAdds a second-line option

Does psilocybin work for OCD? The first randomized trial (2026)

March 2026 · Moreno et al., Journal of Psychopharmacology · Randomized trial, 15 people, double-blind for 4 weeks then single-blind · Small randomized trial: an early signal

What they did

Fifteen people with OCD were randomized to four weekly sessions of high-dose psilocybin, low-dose psilocybin or an active placebo (lorazepam), five people per group. After that, everyone got four more high-dose sessions, and the researchers tracked symptoms for six months.

What they found

Psilocybin reduced OCD symptoms and the placebo didn't. After eight weeks, 73.3% met the usual bar for response (a drop of at least 35% on the Y-BOCS, the standard OCD severity scale) and 40% met the bar for remission. The gains faded but stayed substantial at six months, with no serious side effects reported (Moreno et al., 2026).

What it doesn't show

  • It doesn't show psilocybin works for OCD in general. Fifteen people can't show that.
  • The 73% comes from the second phase, when nobody was blinded and everyone knew they were getting the real thing.
  • It doesn't show psilocybin outside a clinical research setting, with screening and medical monitoring, is safe or useful.

Justin La Rose's take

The 2026 psilocybin trial is a promising first signal, not proof that psilocybin treats OCD: 15 people, five per group in the blinded phase, and a 73% response rate measured after nobody was blinded anymore. I read it the way I read any single small trial. A small trial is a witness, not a verdict. Worth hearing, never the whole case.

Should this change anything for you?

Not yet. Psilocybin isn't an approved OCD treatment, and the IOCDF advises against psychedelic use as treatment outside controlled research. If you're curious, ask about clinical trials. More on how it compares in TMS vs ketamine vs psilocybin for OCD.

Read the study on PubMed

What is the best OCD treatment right now? The BMJ review (2026)

February 2026 · Abramowitz, Abramovitch, McKay & Draffin, BMJ (State of the Art Review) · Review of trials, meta-analyses and guidelines published 2010 to 2025 · Review of the evidence: strong for averages

What they did

Four OCD researchers reviewed fifteen years of trials, meta-analyses and clinical guidelines for adult OCD, from first-line treatments to brain stimulation and psychedelics. People with lived experience of OCD, through the International OCD Foundation, reviewed the draft.

What they found

ERP is still the first-line psychological treatment and SSRIs the first-line medications. Across 30 studies with 1,793 patients, ERP had a large effect compared with placebo and a moderate one compared with other active treatments. About 20% of people don't respond to ERP, about half don't respond to their first SSRI, and doing the practice between sessions, especially in sessions five through nine, predicts how well people do (Abramowitz et al., 2026).

What it doesn't show

  • It isn't independent of the guidelines: the authors are involved with the IOCDF and helped compile its 2025 guidelines.
  • It went to press before the first psilocybin randomized trial was published in March 2026.
  • Averages across trials can't tell you how one person will respond.

Justin La Rose's take

The 2026 BMJ review confirms that fifteen years of OCD research haven't changed the order of treatment: ERP and SSRIs first, everything else after. The finding I'd underline is that doing the practice between sessions, especially in sessions five through nine, predicts how well people do. It's just reps. The work between sessions is the work.

Should this change anything for you?

If you're starting treatment, start with ERP, medication or both. The full ladder is on OCD treatment in 2026.

Read the review in the BMJ

What do the IOCDF's 2025 OCD treatment guidelines recommend?

2025 (guide read October 2026) · International OCD Foundation, OCD Treatment Guide · Clinical guideline: expert synthesis of the evidence · Guideline: strongest for what to do first

What they did

The International OCD Foundation sorted OCD treatments into first-line, second-line, third-line, experimental and treatments to avoid, for adults, children and adolescents.

What they found

First line: ERP, SSRIs and clomipramine. Second line: ACT, cognitive therapy, I-CBT and metacognitive therapy, with some other medications. Third line, for severe treatment-resistant OCD: TMS and deep brain stimulation. Experimental: ketamine and psychedelics. The guide calls the evidence on combining ERP with medication inconclusive, and lists EMDR, hypnotherapy, psychoanalysis and Brainspotting among treatments not recommended on their own (IOCDF).

What it doesn't show

  • A guideline says what has the best evidence on average, not what will work for you.
  • Guidelines lag new studies; it doesn't yet weigh the 2026 psilocybin trial.
  • A treatment missing from first line hasn't been proven useless. EMDR, for example, has a few small trials behind it, including a 55-person pilot where outcomes were comparable to ERP-based CBT (Marsden et al., 2018), but far less evidence overall.

Justin La Rose's take

The IOCDF's 2025 guidelines are the clearest map of OCD treatment we have, and the order they set matters: ERP and SSRIs first, therapies like ACT and I-CBT second, TMS and deep brain stimulation third. Everyone at The OCD Clinic uses ERP, and our approach is informed by all the traditions of psychotherapy, because no two people with OCD need exactly the same thing. Start on the bottom rung. Climb only when that rung, done properly, hasn't been enough.

Should this change anything for you?

If you've been offered something from the avoid list as your main OCD treatment, ask why, and ask about ERP. See treatment-resistant OCD if the first rung hasn't worked.

Read the IOCDF treatment guide

Can beliefs in pregnancy predict postpartum OCD? (2025)

September 2025 · Abramowitz et al., Psychiatry Research · Prospective study, 256 women followed from mid-pregnancy to 6 months after birth · Prospective cohort: shows prediction, not cause

What they did

Researchers measured obsessive beliefs, like how much a thought means or how responsible you are for preventing harm, at 20 weeks of pregnancy. Then they tracked obsessive-compulsive symptoms at six weeks and six months after birth.

What they found

Postpartum obsessive-compulsive symptoms were common: 87.1% reported some at six weeks. Stronger obsessive beliefs during pregnancy predicted more severe postpartum symptoms and a higher chance of an OCD diagnosis at six months, even after accounting for earlier symptoms (Abramowitz et al., 2025).

What it doesn't show

  • It doesn't show that beliefs cause OCD; it shows they predict it.
  • It studied mothers, so it says less about fathers and non-birthing parents.
  • Having some obsessive-compulsive symptoms isn't the same as having OCD.

Justin La Rose's take

The 2025 postpartum study backs up what I see in my practice: in OCD, the intrusive thought isn't the problem, the meaning a person gives it is. Obsessive beliefs measured in pregnancy predicted OCD after birth, which makes those beliefs worth working on early. That's exactly what ERP goes after.

Should this change anything for you?

If you're pregnant and already scared of your own thoughts, that's worth talking about now, not after the baby comes. See postpartum OCD.

Read the study on PubMed

Is I-CBT as good as ERP for OCD? (2024)

October 2024 · Wolf et al., Psychotherapy and Psychosomatics · Multisite randomized trial, 197 people, 20 sessions each · Large randomized trial: good, but one study

What they did

People with OCD in the Netherlands were randomized to 20 sessions of either inference-based CBT (I-CBT), which works on the reasoning behind obsessional doubt without exposure exercises, or standard CBT with ERP, then followed for a year.

What they found

Both groups improved substantially, with no statistically significant difference between them. But the uncertainty around the result was wide enough that the researchers couldn't confirm I-CBT is as good as CBT, so they called it inconclusive. People found I-CBT easier to tolerate (Wolf et al., 2024).

What it doesn't show

  • It doesn't show I-CBT is as good as ERP. That was the question, and the answer was inconclusive.
  • It doesn't show I-CBT is worse, either.
  • It's one trial; the BMJ review notes most I-CBT research so far comes from the people who developed it.

Justin La Rose's take

The 2024 I-CBT trial showed that I-CBT is easier to tolerate than ERP-based CBT but couldn't show it's as effective, and in my practice I use both. I reach for I-CBT more with people who are really uncomfortable doing exposures, because understanding the inferences and the obsessional story someone has built can be the way in. The risk is that a mind that loves to analyze turns the analysis into one more compulsion, so it takes a deft therapist, and the research doesn't settle that question. So I work with the obsessional story, and when it helps, I make that story the object of the exposures.

Should this change anything for you?

ERP stays the core, and I-CBT is a second-line option in the guidelines. In my practice the two work well together: the obsessional story gives the exposures their target, especially when exposure feels impossible at first.

Read the study on PubMed

How strong is each kind of OCD study?

Not every study is worth the same. A guideline or a review of many trials tells you far more than one small trial. Here's how to weigh what you read:

TypeWhat it isGood forWatch out for
Clinical guidelineExperts weigh all the evidence and rank treatmentsWhat to try firstCan lag behind new studies
Review or meta-analysisPools many trials into one answerHow well a treatment works on averageOnly as good as the trials inside it
Large randomized trialHundreds of people, assigned to treatments by chanceWhether one treatment beats anotherStill one study, in one setting
Small randomized trialDozens of people or fewer, assigned by chanceWhether something is worth a bigger trialEasily skewed by chance
Prospective cohortFollows people over time without treating themWhat predicts who develops a problemShows prediction, not cause
Open-label or case seriesEveryone gets the treatment and knows itWhether an idea is worth testingCan't separate the treatment from expectations

How should OCD research guide your own treatment?

OCD research should guide your treatment as a starting point, not a script. Research aggregates data from hundreds or thousands of people into averages. Therapy is more like a research study on one person: you. There are as many ways to get better as there are people, and what you need can change from one season of your life to the next.

So we go into it as scientists. We start with what the evidence says works best for OCD, which is ERP, and we treat the plan as a set of assumptions to test. Every exposure tests what the fear predicts. Every few weeks we test the plan itself: is it working, and how do we know? We stay open to other possibilities, and when your own results say a route isn't working, we consider alternatives, whether that's a different focus, I-CBT, medication or more intensive care.

That's also why progress should be measured, not guessed. Across psychotherapy research, tracking progress and feeding it back into therapy gave a small but real improvement in outcomes (de Jong et al., 2021). If you're choosing a therapist, ask how they'll know if it's working. More on that in how to know if therapy is working.

Which older OCD studies still matter?

The OCD studies that still matter are the ones every new trial gets measured against. These are the findings I lean on most:

  • ERP works across dozens of trials. A meta-analysis of 37 randomized trials found large effects for CBT including ERP compared with waitlist and placebo (Öst et al., 2015).
  • ERP gains tend to hold. Twelve weeks after treatment stopped, 12% relapsed after ERP compared with 45% after clomipramine alone (Simpson et al., 2004).
  • Exposure works by learning, not by calming down. Exposure works through expectancy violation: learning the feared outcome doesn't arrive, or that you can carry the uncertainty (Craske et al., 2014).
  • Family reassurance keeps OCD going. Family accommodation was linked with more severe OCD across 41 studies (Wu et al., 2016).
  • Postpartum harm thoughts are common. Of 388 new mothers, 44.4% reported unwanted thoughts of intentionally harming their infant, and those thoughts weren't linked to more reported aggression in that sample (Fairbrother et al., 2022).
  • Face-to-face CBT still leads. Across 25 trials, face-to-face CBT outperformed therapist-guided internet CBT, and all formats beat waiting lists (Zhang et al., 2023).

What's next in OCD research?

What's next in OCD research is mostly bigger, better versions of the studies above. The BMJ review flags these open questions (Abramowitz et al., 2026):

  • Larger randomized trials of psilocybin, including psilocybin paired with psychotherapy
  • Whether ERP designed around inhibitory learning beats standard ERP in OCD-specific trials
  • Whether bringing a partner or family member into ERP improves results
  • Who responds best to which treatment, so care can be matched from the start

When those land, they'll be here, with the same five questions. For the full picture of what works today, see OCD treatment in 2026.

FAQ

What Else Do People Ask About OCD Research?

The newest treatments being studied for OCD are psilocybin, ketamine and refined forms of brain stimulation like TMS. Psilocybin and ketamine are still experimental. The best-supported treatments are still ERP and SSRI medication.

No. Psilocybin is not an approved treatment for OCD. The first randomized trial, published in 2026, had 15 participants, and its authors say larger trials are needed. The IOCDF advises against psychedelic use as treatment outside controlled research.

There is no single cure, but OCD is treatable. Many people reach the point where OCD no longer runs their life, while intrusive thoughts still show up sometimes. ERP and SSRIs remain the treatments with the strongest evidence.

Look at how many people were studied, whether they were randomly assigned, whether anyone was blinded, and whether results held at follow-up. A small, unblinded study is an early signal. A review of many trials is much stronger.

Want Help Making Sense Of It?

Book a free 20-minute consultation. No commitment, no pressure. Just a conversation about where you are and how we can help.

Book Free Consultation
Online across Ontario · Toronto · In-Person and Video Sessions · Evening & Weekend Available · Many extended health plans cover Registered Psychotherapists

How we researched this page

Each entry on this page is checked against the original paper's abstract on PubMed or the source's own website, and the numbers are reported as the authors give them. Entries cover peer-reviewed studies and clinical guidelines on adult OCD published from 2024 onward, with older foundational studies listed separately. Small trials and single studies are labelled as such, and none of these findings predicts how any one person will respond to treatment.

A note on my perspective

I run a private clinic that treats OCD with ERP, and I'm paid when people book with us. Read what I write with that in mind. Public and lower-cost routes exist too, and the Ontario OCD treatment guide lays them out next to ours.

About the author

Justin La Rose, RP, M.Psy is a Registered Psychotherapist with the College of Registered Psychotherapists of Ontario (CRPO) and Clinical Director of The OCD Clinic in Toronto. He has worked in mental health for more than 25 years and treats OCD with Exposure and Response Prevention (ERP). You can confirm his registration on the CRPO public register.

Client stories on this site are composites, with names and identifying details changed. How we write client stories

How to cite this page

La Rose, J. (2026). OCD Research Updates: What's New and What It Means. The OCD Clinic. https://theocdclinic.ca/ocd-research