Think Differently. Solve the Impossible.

MDMA, known recreationally as Ecstasy or Molly, has been studied as part of psychotherapy for post-traumatic stress disorder (PTSD).

I remember the warnings about Ecstasy as a recreational drug in the 1990s and 2000s. The message then was simple: Ecstasy is bad for you. Don't take it.

Hearing about MDMA being used with psychotherapy I had questions:

• If it's dangerous recreationally, what is it ok to use with psychotherapy?
• How does the treatment work?
• Is hallucination required as part of  the treatment, or am I confusing MDMA with drugs such as LSD and psilocybin?
• If the treatment is useful, why isn't it legal?

I invited Jill Sitnick, author of Rescuing Jill, to be a guest on my show to have a conversation. Jill underwent MDMA-assisted psychotherapy after being diagnosed with PTSD. Jill isn't a physician or medical researcher. What she brought was firsthand experience with the treatment and knowledge she's accumulated from her experience.

She corrected some of my assumptions. At one point I asked whether MDMA could treat anxiety, depression, and other mental-health problems. She didn't want me to become, as she put it, “medicine blind” by suggesting MDMA can work for everything. She preferred to stay with what had been studied in clinical trials for PTSD.

We might think, Ecstasy is a dangerous street drug, and therefore MDMA-assisted therapy must be dangerous. We can make the same mistake in the opposite direction, with conclusions such as: MDMA helped Jill's PTSD, therefore MDMA is a safe and effective treatment for mental-health problems.

Neither conclusion follows. During our conversation, Jill added context. Dose and the circumstances under which the drug is taken matter. Taking MDMA wasn't the entirety of Jill's treatment. Before we get into what MDMA-assisted therapy is, let's start with what MDMA is.
 

What Is MDMA?

MDMA stands for 3,4-methylenedioxymethamphetamine.

The National Institute on Drug Abuse describes MDMA as a synthetic drug that acts as both a stimulant and psychedelic and can produce increased energy, pleasure, emotional warmth, and altered sensory and time perception.

My assumption wasn't wrong. Ecstasy and Molly are names associated with MDMA. Every experience taking Ecstasy is not equivalent to MDMA-assisted psychotherapy.

Same Drug. Same Risk?

Jill suggested the dose may be what's different. She compared it with other medication. Taking the prescribed amount of something isn't the same as taking many times that amount. When we're talking about a prescription drug, we don't say, “This medication can cause serious harm if you take ten times the prescribed amount, therefore the medication has no legitimate medical use.” We ask what happens at the dose at which it's intended to be used. Why wouldn't we ask the same question about MDMA?

That doesn't establish that MDMA as safe. It establishes that risk depends partly on exposure: Dose. Frequency. Purity. Other substances being taken at the same time. The health of the person taking it. The environment in which it's taken.

When I hear “Ecstasy is dangerous,” the questions now are:

Under what conditions?
Are those the same conditions in MDMA-assisted psychotherapy?

That's not an argument for MDMA. It's an argument for comparing things.
 

What You Might Get When You Buy Molly

There's a problem with making a direct comparison between recreational Molly use and MDMA used in research and controlled therapy. A substance sold illegally as Molly isn't a controlled pharmaceutical product whose composition is known. That introduces a variable that isn't inherent to MDMA:

What did the person take?

If someone experiences an adverse effect after taking something sold as Molly, we need to know whether it contained MDMA, how much it contained, and what else may have been present before deciding what the event tells us about MDMA.

That doesn't make recreational MDMA safe. It means “someone took Molly” isn't a complete description of an exposure. The distinction becomes important if we're trying to use recreational-drug experiences to judge a controlled medical treatment.
 

What MDMA Does in Therapy

This is where Jill's experience became more useful than a general description of MDMA. Jill described MDMA as, “absolutely beautiful for PTSD because it calms the body down.” That's her description of what she experienced, not a medical conclusion I'm asking the reader to accept. It helped me understand what and why she believed MDMA contributed to her treatment.

She described being able to revisit painful childhood memories without experiencing the same overwhelming physical fear she had previously associated with them. She wasn't trying to forget what happened to her. She wasn't trying to hallucinate. The memories remained. Her response to them changed.

Jill described being able to reconsider beliefs created by an abused child's mind from the perspective of an adult, without her body reacting with the fear it had carried for decades.

That question: is MDMA changing the memory or changing the conditions under which she could examine the memory? Jill's experience suggests the latter.

That becomes important when we hear the simplified statement: MDMA treats PTSD. That's compressing what happened too far. MDMA's role in Jill's treatment wasn't to eliminate the traumatic memories, rather it allowed her to approach those memories differently, in a manner by which psychotherapy could accomplish something previously too difficult or impossible.

That leads to another question.

If that's what MDMA was contributing, was hallucination part of the objective?
 

MDMA Isn't the Same Experience as Magic Mushrooms

Jill also used psilocybin, commonly associated with magic mushrooms. She described that experience differently. For her, mushrooms could be more visual. MDMA, by comparison, produced a calmer state in which she could engage with what was happening without feeling as though she were dramatically altered.

I asked about LSD. She had only microdosed LSD and said she couldn't personally describe what a full LSD experience is like. That's worth noting, as we tend to lump these substances together as psychedelics, then transfer characteristics from one to another.

Whatever characteristic we're discussing can become attached to the category rather than the drug, dose, and circumstances.
 

What the Treatment Involves

An assumption about MDMA-assisted therapy is that it only involves taking MDMA. Jill didn't describe going to a doctor, taking MDMA, and having her PTSD disappear. Her treatment process took roughly a year and included three MDMA-assisted sessions.

Before the first session, there were months of talk therapy and preparation. During the MDMA sessions she worked with two people, a male and female therapist/guide team, following the MAPS protocol.

The physical environment was intentionally designed to feel comfortable. Jill described a room with artwork, blankets, seating areas, and a mattress. The mattress provided a safe place to sit or move around if difficult material emerged.

Before taking the MDMA, they discussed her intentions for the session. Jill's was straightforward: Fear. She wanted to understand and change the fear response that had been dominating her life.

During the session, the people with her weren't simply watching for medical complications. Jill described them as “holding space,” being present while she worked through whatever emerged.

When her first session ended, she thought it hadn't worked. On the ride home, childhood memories began coming back to her, and she began interpreting some of those memories differently.

We'll come back to what changed.

The important point is that if someone says, “MDMA treated Jill's PTSD,” that's an enormous amount of context compressed into one sentence. What Jill underwent was MDMA-assisted psychotherapy, not simply MDMA.
 

Adding Variables

Jill's third session added a variable to discussion about MDMA-assisted therapy. 

Her guide and therapist were concerned that there was traumatic material she couldn't reach using only MDMA. Jill said they discussed whether she would be willing to add a small amount of psilocybin during the MDMA session if necessary.

She agreed.

Why combine them? The following come to mind:

Wouldn't more MDMA accomplish the same thing? 
Why wouldn't psilocybin alone work as good? 
What are the two drugs doing differently?

Jill couldn't answer that, chemically. She told me she was looking forward to research that could explain the mechanisms. She told me what she experienced as a patient. An experience can tell us that something happened without telling us why.

Adding psilocybin introduces another variable. If something changed during Jill's third session, how much can we attribute to the MDMA, how much to the psilocybin, and how much to the combination? We don't know.

Jill's particular treatment shouldn't be treated as a description of a standard MDMA-assisted psychotherapy protocol, as her third session involved circumstances specific to her treatment.

What Part of the Treatment Is Doing the Treating?

Jill emphasized that the circumstances under which the drug is taken matter.

What if the circumstances aren't merely there to make taking MDMA safer? What if they're part of what makes the treatment work?

MDMA-assisted psychotherapy doesn't involve simply administering a drug. There's preparation and a particular environment. There are therapists present during the session, and interactions with those therapists in integration sessions afterward.

When we ask, does MDMA work?

That's not what was tested.

The studies compared psychotherapy + MDMA with psychotherapy + placebo. All the participants received preparatory sessions, extended dosing sessions, and subsequent therapeutic sessions.

Both groups improved.

In a 2023 Phase 3 study, 71.2% of participants evaluated in the MDMA-assisted therapy group no longer met the diagnostic criteria for PTSD at the study's conclusion, compared with 47.6% in the placebo-with-therapy group.

That suggests that adding MDMA contributed something to the treatment. If psychotherapy were doing everything and MDMA were doing nothing, we wouldn't expect the MDMA group to outperform the placebo group.

There's a missing comparison:

MDMA without psychotherapy.

Interestingly, the FDA noticed this too. The FDA review noted that the studies demonstrated a difference between the MDMA and placebo groups, suggesting an effect attributable to MDMA. It did not establish whether the psychotherapy independently contributed to the treatment effect of the combination.

That changes the question.

Instead of asking, Does MDMA treat PTSD?

The question is:

What does MDMA contribute to a therapeutic process that is already doing something?

And: Would MDMA produce the same results without that process?

The studies we're discussing weren't designed to answer that question. That's why MDMA-assisted psychotherapy may be a more important phrase than it initially appears.

The assisted part isn't an accessory to the drug.

It's part of the treatment.

The FDA isn't merely asking, “Is this drug safe and effective?” It has to evaluate a treatment in which the drug, therapists, setting, preparation, expectations, and subsequent integration are difficult to completely disentangle.
 

If It Works, Why Isn't It Approved?

At the time of our conversation, Jill was optimistic that FDA approval was coming.

It didn't.

In August 2024, the FDA declined to approve Lykos Therapeutics' application for MDMA-assisted therapy for PTSD and requested an additional Phase 3 trial.

Why wouldn't the FDA approve it?

One possible conclusion is, it must not work. Another is, the FDA won't approve it because of the stigma surrounding psychedelics. Neither is sufficient.

The FDA's decision wasn't simply, MDMA doesn't work. It was closer to, the evidence submitted wasn't sufficient to establish that this particular treatment should be approved as proposed.

FDA approval also would not have answered my original question about whether the treatment was legal. MDMA is federally controlled, therefore, FDA approval of a medical use would have implications for its scheduling under federal controlled-substance law. State law and professional regulation can add another layer.

FDA approval wouldn't necessarily mean every therapist in every state could offer MDMA-assisted psychotherapy. That's an example of how a regulatory process gets compressed into one word: Legal.
 

Is “Psychotherapy” a Single Variable?

We've been treating psychotherapy as a single variable. Researchers have been dealing with this problem in for a long time. They try to create something resembling the psychotherapy equivalent of a placebo, something considerably harder than making a sugar pill. These are called attention controls, supportive controls, nonspecific-factor controls, or structurally equivalent controls. The idea is to hold as many things constant as possible:

Time with the therapist.
Number and length of sessions.
Attention from another person.
Expectation of improvement.
The opportunity to talk.

Researchers change or remove a particular therapeutic technique they're trying to evaluate.

Researchers can break psychotherapy down further using what are called component or dismantling studies. Start with the complete therapy and remove one component, or add a particular component to a simpler intervention, and see what changes.

That suggests an MDMA experiment could become considerably more sophisticated than MDMA + psychotherapy versus placebo + psychotherapy.

You might ask:

MDMA + full psychotherapy
Placebo + full psychotherapy
MDMA + nonspecific/supportive therapist interaction
Placebo + nonspecific/supportive therapist interaction

That's a factorial design.

Interestingly, the FDA suggested essentially this approach. Its guidance for psychedelic-drug studies discusses factorial designs as a way of evaluating the separate contributions of the drug and psychotherapy. The FDA says sponsors should justify including psychotherapy. This isn't an experiment I thought of. The FDA has identified the same problem.

But there's a complication.
 

What Is the Active Ingredient in Psychotherapy?

Suppose the psychotherapy consists of time with another person + empathy + trust + expectation + talking about the trauma + confronting memories + reframing them + particular therapeutic techniques.

Which part is the treatment? 

If we try to create a placebo psychotherapy consisting of time + empathy + trust + expectation + talking, we may accidentally create something therapeutic.

That's a known problem in psychotherapy research. The things we might dismiss as background conditions—the relationship with the therapist, expectations, communication, support, engagement—may contribute to the outcome.

Psychotherapy isn't a single variable.

That brings us back to MDMA.

With MDMA + psychotherapy vs. placebo + psychotherapy, 
we need to break the second half of the equation apart.

What does the preparation contribute? What does trust in a therapist contribute? What happens during the MDMA session? Does confronting the traumatic memory matter? Does the therapist need to actively intervene? What does the environment contribute? How important is integration afterward?

Perhaps the most interesting question:

Which of those things does MDMA enhance?

If the proposed purpose of MDMA isn't to eliminate PTSD pharmacologically, rather it's to make some part of the therapeutic process work differently or more effectively, that's what we  need to identify.
 

Can a Scalpel Fix It?

Imagine a surgeon saying, I can perform this operation with a dull instrument. If you give me a properly designed sharp surgical instrument, I can perform it more effectively.

We wouldn't test the usefulness of the scalpel by handing it to somebody in an empty room and asking, does the scalpel cure the disease? It doesn't. We would not conclude, surgery is what fixed the patient, therefore the scalpel contributed nothing.

We'd ask, does the surgeon perform the procedure better with the instrument than without it, and does the additional benefit justify the instrument's risks?

That resembles the MDMA experiment:

Psychotherapy + placebo → improvement
Psychotherapy + MDMA → greater improvement

Maybe MDMA is closer to a tool that enables the therapist and patient to do something more effectively. Jill's description fits that hypothesis. She didn't describe MDMA as going into her brain and erasing PTSD. She described being able to encounter memories without the overwhelming fear response that previously prevented her from dealing with them differently.

If that's part of the mechanism, then asking, does MDMA work without psychotherapy, could be somewhat like asking: Does a scalpel work without surgery?

The answer would be, No. Why would we expect it to?

That wouldn't demonstrate that the scalpel doesn't work. It would tell us we've misunderstood what the tool is for.
 

Where the analogy becomes useful

A sharp knife can cause harm.

We don't prohibit surgeons from using scalpels because knives are dangerous when used improperly. We control who uses it, where it's used, what it's used for, and under what circumstances.

That's close to Jill's point - the circumstances under which the drug is taken matter.

The question isn't, is MDMA dangerous?

It can be.

The question isn't, does MDMA cure PTSD (by itself)?

(My instinct is to say no, it does not. People have been taking Ecstasy recreationally for decades. If taking MDMA by itself reliably cured PTSD, wouldn't we expect that effect to have become apparent?)

I can't make that conclusion from recreational use. I don't know which recreational users had PTSD before taking MDMA, what they took, at what dose or purity, under what circumstances, or what happened to their PTSD symptoms afterward. The clinical trials we're discussing didn't test MDMA without psychotherapy.

The accurate answer is:

We don't know.

We do know the treatment being proposed isn't simply MDMA. It's MDMA used as part of a therapeutic process. The relevant question for the proposed treatment might be:

Does MDMA provide a useful therapeutic capability when used by appropriately trained people, under defined conditions, as part of a treatment that produces better outcomes than the treatment without it—and do those benefits justify its additional risks?

That's a precise question.

Interestingly, FDA's 2026 guidance suggests the agency is grappling with exactly this interaction problem: it says psychedelic development frequently combines the investigational drug with psychological support or psychotherapy, which complicates determining effectiveness and future labeling.

Maybe We're Asking the Wrong Question

The obvious question is:

Does MDMA work?

That's not what was tested.

The studies compared psychotherapy plus MDMA with psychotherapy plus placebo.

Both groups improved.

The MDMA group improved more.

That gives us evidence that adding MDMA contributed something. If psychotherapy were doing everything and MDMA were doing nothing, we wouldn't expect the MDMA group to consistently outperform the placebo group.

There's a missing comparison:

MDMA without psychotherapy.

The FDA noticed this too. In its review, the FDA stated that there were no data on the efficacy of MDMA without psychotherapy. There also weren't data comparing the psychotherapy component with no treatment.

The experiment looks something like this:

We know:  Psychotherapy + placebo → improvement

and: Psychotherapy + MDMA → greater improvement.

We don't know: MDMA without psychotherapy → ?

Interestingly, the FDA noted that the development program essentially assumed psychotherapy was necessary for achieving the therapeutic response.

At first, that sounds like a weakness in the evidence.

Maybe it is.

If we're trying to determine whether MDMA itself treats PTSD, wouldn't we want to know what happens when someone with PTSD receives MDMA without the psychotherapy? That made me wonder whether we're asking the question backward.
 

What If the Psychotherapy Isn't a Confounding Variable?

Jill repeatedly emphasized that the circumstances under which MDMA is taken matter. I thought of that primarily as a safety issue. Taking a known dose of pharmaceutical MDMA under professional supervision isn't the same exposure as taking an unknown quantity of something sold as Molly at a club.

Perhaps the circumstances aren't merely there to make taking MDMA safer. Maybe they're integral to what makes the treatment work.

Think about what Jill described.

She didn't simply take MDMA.

She prepared with therapists beforehand. She entered the session with an intention, and was in a controlled environment. Two people were there with her. They helped her work through what emerged. She revisited traumatic memories while experiencing them differently. She continued working with those memories afterward.

If that's the treatment, separating MDMA from the psychotherapy may tell us something scientifically interesting without necessarily answering the clinical question we're trying to solve.

Perhaps MDMA isn't the treatment.

Perhaps MDMA is something that makes a particular kind of treatment possible or more effective.

Jill's description suggests that MDMA changed the conditions under which she could encounter traumatic memories. The therapists then helped her work within those conditions. The integration afterward helped her reconsider and reinforce what emerged.

If that's how the treatment works, asking:

Does MDMA treat PTSD without psychotherapy? may be a little like asking, does anesthesia make surgery successful without the surgeon? Anesthesia by itself doesn't repair your hip. That doesn't mean anesthesia isn't doing something important during hip surgery. The question is what role each component plays in producing the outcome.
 

What Did the Studies Establish?

Both groups received the therapeutic intervention.

One received MDMA. The other received placebo.

The MDMA group improved more.

Unless that difference is explained by a problem with the experiment, the results suggest MDMA added something to the therapeutic process.

The treatment being proposed: MDMA-assisted psychotherapy.

Not MDMA by itself.

Does adding MDMA to an appropriately controlled therapeutic process improve outcomes enough to justify the additional risks?

Studies suggest that it might.

That doesn't answer the FDA's problem. MDMA introduces another variable that's much harder to control:

People can feel it.
 

The Problem With Studying a Drug You Can Feel

Clinical trials often use blinding. Participants don't know whether they're receiving the drug being studied or a placebo. MDMA produces noticeable effects. That creates what the FDA described as functional unblinding.

If I participate in a trial and experience the characteristic effects of MDMA, there's a good chance I can guess which group I'm in. If I take the placebo and experience none of them, I may also be able to guess.

Expectations can influence reported outcomes. That doesn't mean a treatment didn't work. It means one of the methods we rely upon to separate a drug effect from expectations is difficult.
 

What Concerned the FDA

The FDA specifically identified functional unblinding as an interpretability problem. Because people can generally feel MDMA's effects, participants, and potentially therapists, could often infer who received MDMA and who received placebo.

The advisory committee said expectation bias and functional unblinding “may have played a role in the efficacy results.” There were other concerns involving the evidence and safety information as well.

Jill raised a different question.

She wondered whether decades of the war on drugs continue to influence how we think about treatments involving drugs such as MDMA. 

That's her opinion.It's not the FDA's stated reason for declining approval. I haven't found evidence that the FDA rejected the application because it feared medical approval would lead to broader legalization of MDMA.

The FDA could approve MDMA for use under particular therapeutic conditions, without making recreational MDMA legal. MDMA's status as a controlled substance creates another regulatory layer beyond FDA approval.

I don't think the question should be, is the FDA refusing to approve MDMA because it doesn't want MDMA legalized? A better question is:

Does MDMA's history as an illegal recreational drug affect how we evaluate its potential medical use?

A drug called Ecstasy arrives with cultural baggage that a newly developed pharmaceutical doesn't. An  assumption might be: Ecstasy is a dangerous recreational drug. How did we get from there to giving MDMA to people with PTSD?

That doesn't mean the FDA's concerns are stigma disguised as science. The concerns about safety, study design, functional unblinding, therapist conduct, and the quality of the evidence need to stand on their own.

It  makes Jill's question interesting:

How much of the resistance to psychedelic-assisted treatments comes from unresolved scientific and safety questions, and how much, if any, comes from the cultural and legal history of the drugs themselves?

I don't know.

I don't think Jill knows.
 

An interesting paradox

The thing that makes MDMA-assisted therapy difficult to evaluate may also be the thing that makes it work. The FDA understandably wants to know what contribution belongs to MDMA, and what contribution belongs to psychotherapy.

Jill's experience suggests those things may interact. MDMA changes the conditions under which she encounters the memory. The therapists help her work with what emerges. The integration afterward helps her continue processing it.

Perhaps we're trying to isolate components of something whose value comes from the interaction between the components.

That doesn't mean we shouldn't try to separate them scientifically.

We should.

We need to be careful about the conclusion we draw.

“We don't know exactly how much of the effect belongs to each component” is not the same statement as, “We don't know whether the combined treatment works.”

Neither statement answers the regulatory question:

“Do we know enough about the benefits and risks to approve it?”
 

Why Not Approve It Under Specific Conditions?

If FDA understands the problem, why didn't it approve MDMA under restricted conditions?

FDA's position in 2024 was essentially:

We aren't yet convinced that the evidence establishes a favorable benefit-risk profile, even under the proposed conditions.

That's different from saying:

We believe it works, but we can't figure out how to control its use.

The FDA's Complete Response Letter says the application could not be approved in its present form and identified several deficiencies. The FDA requested another adequate and well-controlled study. There were overlapping problems.

The efficacy signal was positive. The FDA questioned how confidently it could interpret it This is where functional unblinding matters. Participants could frequently tell whether they received MDMA. That changes expectations, reporting, therapist behavior and even how outcomes are assessed.

Although:

MDMA + psychotherapy performed better than placebo + psychotherapy,

The FDA wasn't asking whether the numbers were statistically different. It was asking:

How much confidence should we have that the measured difference represents the pharmacological treatment effect rather than biases introduced by the study design?

That's a higher bar.
 

FDA also thought the safety database was insufficient

The Complete Response Letter raised concerns about cardiovascular effects and requested additional characterization of abuse-related effects and other safety issues. FDA wanted better information about blood pressure effects and specifically requested ambulatory blood-pressure monitoring in a new study.

A restricted-use program can manage a known risk; it can't substitute for knowing what the risk actually is.

There's a difference between:

We know this causes X risk, so patients must be monitored for X

and: We don't yet have adequate evidence to characterize X.

A restricted-use condition solves the first problem better than the second.

There were data-integrity and study-conduct concerns

FDA's letter raised concerns about protocol deviations, adverse-event reporting and aspects of study conduct. The advisory process had also scrutinized therapist conduct and the potential effects of prior MDMA experience among participants.

That's harder to repair with: Only allow trained therapists to administer it.

Interestingly, there's a legitimate controversy. MAPS points out that FDA had agreed in 2017 through a Special Protocol Assessment that the Phase 3 protocol design, endpoints and planned analyses were acceptable.

MAPS argues that functional unblinding was discussed with FDA during that process, yet became an important objection during the 2024 review. That's MAPS's characterization of the history, not an FDA admission that it changed the rules. If the problem is partly the evidence itself, that left me with another question:
 

What Have They Done Differently?

Given the status of FDA approval, have the FDA or those seeking approval done anything differently in the way they design the studies to answer these questions and move toward approval—or determine that it shouldn't be approved?

As of August 2026, the answer is Yes. Not how I expected.

They didn't simply redo the Phase 3 Trial. In 2024, the FDA declined to approve Lykos Therapeutics' application and requested an additional Phase 3 trial. Two years later, that additional Phase 3 trial apparently hasn't happened.

In August 2026, MAPS reported that Resilient Pharmaceuticals, formerly Lykos Therapeutics, had resubmitted the MDMA/PTSD New Drug Application to the FDA without conducting a new Phase 3 study.

They apparently aren't responding by saying:

You're right. We'll run the experiment again differently and answer all of these questions. They're asking FDA to reconsider the application without the additional Phase 3 trial FDA originally requested.

I don't know everything submitted or how it has attempted to address each of FDA's concerns. MAPS says the FDA's 2024 response recommended additional research involving durability of response, safety characterization, and reducing potential bias.

For this application, we haven't yet gotten the experiment I was imagining:

psychotherapy + MDMA

versus

psychotherapy + placebo

versus

MDMA without psychotherapy

versus

psychotherapy alone or no treatment.
 

The FDA Changed the Playbook

In July 2026, the FDA finalized its first guidance specifically addressing clinical investigations of psychedelic drugs. The FDA says psychedelic drugs present “unique challenges” in designing clinical trials and specifically addresses study design, data collection, patient monitoring, and how to conduct adequate and well-controlled studies.

FDA: Psychedelic Drugs — Considerations for Clinical Investigations

The FDA isn't saying:

Psychedelics are too difficult to study.

It's saying:

They present unusual experimental problems. Here's how we think those problems should be addressed.

Some of those problems are the ones we've been talking about. These aren't new concerns. The FDA first issued draft psychedelic-trial guidance in 2023, specifically identifying psychotherapy, safety monitoring, dose-response, durability, and the difficulty of designing adequate and well-controlled trials as issues researchers needed to address. The July 2026 guidance finalizes that earlier work.

The FDA appears to be trying to improve the experimental framework.
 

Researchers Are Changing the Therapy Variable

One of the problems we've identified is that “psychotherapy” can be a significant variable. What are the therapists doing? Could the MDMA simply be enabling whatever the therapists are doing?

There are studies attempting to make that variable more specific.

A current VA-affiliated study is testing MDMA with Massed Prolonged Exposure, an established form of PTSD psychotherapy. Instead of MDMA plus a broadly defined supportive therapeutic process, the researchers are pairing MDMA with a specific psychotherapy protocol.

A larger, randomized, placebo-controlled study is designed to enroll 95 participants and explicitly asks whether Prolonged Exposure + MDMA produces a greater reduction in PTSD symptoms than Prolonged Exposure + placebo. It also asks whether the combination improves durability and examines possible mechanisms involving fear extinction and threat response.

Maybe We Don't Need the Missing Cell

This brought me back to the question:

Does MDMA work?

Maybe that's too vague. Suppose MDMA without psychotherapy produces little or no lasting improvement in PTSD. Would that mean MDMA doesn't work?

If the treatment being proposed is MDMA-assisted psychotherapy, the clinically relevant question may not be whether MDMA independently treats PTSD.

It may be:

Does adding MDMA to an appropriately controlled therapeutic process improve outcomes enough to justify its additional risks?

We have some evidence it does. That leaves another problem:

How confident are we that the difference was actually caused by MDMA?
 

We Still Have the Blinding Problem

One way researchers have tried to address that problem is with an active placebo. Instead of giving the control group something that produces no noticeable effect, researchers can give them something that produces enough of an effect that they aren't immediately certain which group they're in.

Earlier MDMA studies experimented with low-dose MDMA as an active control.

That doesn't eliminate functional unblinding. It demonstrates that researchers have ways of designing experiments intended to make the comparison more meaningful.

That gets us closer to the experiment I want to see.

Keep the psychotherapy as consistent as reasonably possible, then make it harder for the patient and therapist to know who received the therapeutic dose of MDMA.
 

There's an Interesting Regulatory Problem

The problem is more complicated than designing a better experiment. 

If the treatment's effectiveness depends upon:

MDMA + preparation + trained therapists + setting + therapeutic interaction + integration

what is the FDA approving?

The FDA regulates the drug.

It doesn't regulate the practice of psychotherapy. Yet the evidence supporting the drug may depend upon the therapeutic system surrounding it.

You aren't merely evaluating a molecule. You're evaluating a molecule whose demonstrated benefit may depend upon a surrounding system that the drug regulator doesn't completely control.
 

That creates another problem.

Suppose the evidence establishes that:

MDMA + Therapy A works.

What happens when a clinician uses a different one such as Therapy B?

Does that require another FDA approval?

Generally, physicians can use an approved drug off-label when they determine it's medically appropriate. FDA regulates drugs, not the practice of psychotherapy.

That creates a problem.

If the psychotherapy is partly responsible for the outcome, changing the psychotherapy isn't equivalent to a physician making a minor variation in how an otherwise established treatment is delivered.

At some point we may have changed one of the variables responsible for making the treatment work. Make the FDA-approved protocol extremely specific and we constrain clinical judgment and move FDA toward regulating the practice of psychotherapy. Make it too broad and we risk assuming that MDMA + Therapy B works because MDMA + Therapy A worked in a clinical trial.

There's another question researchers need to answer:

Which parts of MDMA-assisted therapy are essential to producing the result, and which can be changed without changing the result?

That's what determines whether the treatment demonstrated in a clinical trial can survive contact with medical practice.

The Latest Development

The story is no longer, FDA rejected MDMA-assisted psychotherapy and requested another Phase 3 trial. As of August 2026, it's: FDA requested another Phase 3 trial. The sponsor has reportedly resubmitted the application without conducting one.

Meanwhile, the FDA has finalized guidance specifically intended to improve how psychedelic-drug trials are designed, and researchers are experimenting with better-defined therapeutic protocols.

This raises three different questions:

Does the complete treatment work? Existing trials provide evidence that MDMA plus psychotherapy produces greater improvement than placebo plus psychotherapy.

What does MDMA itself contribute? The difference between the groups suggests it contributes something. Functional unblinding and the interaction with psychotherapy make the size and nature of that contribution harder to isolate.

Do we know enough about the benefits, risks, and conditions necessary to produce the result to approve it? That's the regulatory question.

And an interesting paradox:

The thing that makes MDMA-assisted psychotherapy difficult to evaluate may also be the thing that makes it work. The drug and the therapeutic process may not be independent variables in the real world. They may be interacting parts of the treatment. If that's true, trying to determine whether MDMA works without psychotherapy is scientifically interesting.
 

Jill Encountered This Problem

This may sound like a hypothetical regulatory problem. It isn't.

Jill's treatment provides an example. Jill underwent three MDMA-assisted journeys. Before her third, her guide and therapist were concerned that there was traumatic material she still couldn't reach. They discussed beforehand whether she would be willing to add a small amount of psilocybin if the MDMA session wasn't getting her there.

She agreed.

They added it.

Jill associates that third journey with an important breakthrough in her treatment.

Now I have another problem.

Jill's treatment isn't the treatment we've been talking about approving.

The clinical evidence being evaluated concerns MDMA administered within a particular therapeutic process. Jill's treatment was modified, based on what her therapist and guide believed she needed.They added another psychoactive drug. From the perspective of a controlled experiment, that's another variable. From the perspective of treating an individual patient, it may look like clinical judgment.

That's where this gets complicated.

What happens when the approved protocol doesn't work for a particular person? A clinician may want to modify it. Different psychotherapy. A different dose. Or, as happened in Jill's case, another substance.At what point have we changed the treatment enough that the clinical trials no longer tell us whether what we're doing is safe or effective?

Jill's experience doesn't establish that adding psilocybin made the treatment work. We don't have the counterfactual Jill who went through the same third session without it.

Maybe the psilocybin was responsible for the breakthrough.
Maybe the first two MDMA sessions made the third possible.

Maybe the psychotherapy did.
Maybe it was the accumulation of a year of work.

We don't know.

Clinical trials try to eliminate variables so we can determine what causes an outcome.
Clinical practice adds variables when the patient isn't responding.

If psychedelic-assisted therapy becomes ordinary clinical practice, researchers and regulators may have to determine not only whether the standardized treatment works, but how much clinicians can change it before they're no longer delivering the treatment that was actually studied.
 

This Is No Longer an MDMA Problem

There's another reason the FDA may have incentive to figure this out now, and may be taking more time to do so. Other psychedelic drugs are being investigated for medical uses, including psilocybin and LSD, and some of them create many of the same problems.

How do you blind a clinical trial when the patient can tell something dramatic is happening?

How do you separate the pharmacological effect of the drug from expectations created by the experience?

What happens when the drug is administered with psychotherapy or another psychological intervention that may itself contribute to the result?

How do you determine whether the benefit comes from the drug, the experience produced by the drug, the therapy surrounding it—or the interaction among all three?

These questions aren't unique to MDMA.

The FDA created guidance specifically for psychedelic-drug development. The agency says interest in the therapeutic potential of psychedelic drugs has been increasing and that these drugs present “unique challenges” for designing studies capable of producing interpretable results.

That changes the incentive to solve the problem. When MDMA was one unusual drug-development program, perhaps these questions could be dealt with largely within that application.

What happens when there are multiple drugs, multiple sponsors, multiple psychiatric conditions, and multiple clinical trials confronting variations of the same problem?

Now the FDA needs something more general:

How should this entire class of treatments be evaluated?

That's what seems to be happening.

The FDA issued its first draft guidance specifically for psychedelic clinical trials in 2023 and finalized that guidance in July 2026. It now maintains resources specifically devoted to psychedelic-drug development and has scheduled a public hearing on the potential future therapeutic use of psychedelic drugs in supervised and supportive settings.

Perhaps what has changed isn't the evidence for MDMA.

The problem has gotten bigger.

If MDMA, psilocybin, LSD, and future drugs continue moving through development, the FDA can't indefinitely treat the unusual characteristics of psychedelic-assisted therapy as an exception.

It needs a framework for evaluating them.

That may ultimately help answer some of the questions the MDMA application exposed.
 

Is That the Whole Explanation?

Let's assume, pretend for a moment, we don't trust FDA. We can explore this without turning it into a conspiracy argument. The thought exercise is: If FDA's stated reasons aren't the whole explanation, what other incentives or institutional forces could produce the same decision?

There are several plausible hypotheses. I don't have evidence establishing that any of them caused the rejection.

• Asymmetric risk. Approving MDMA and later discovering serious problems could be a bigger institutional failure for FDA than delaying a treatment that proves beneficial. Regulators may therefore rationally and excessively favor false negatives over false positives.
• First-mover risk. This wasn't just another antidepressant. It would have been the first FDA-approved psychedelic-assisted treatment of this kind. The first approval establishes precedents involving therapists, setting, monitoring, REMS-type controls, labeling and future psychedelic applications. FDA may therefore demand more certainty from the first case than it might once a regulatory pathway exists.
• The drug's history matters culturally. MDMA arrives carrying decades of association with Ecstasy, Schedule I status and the war on drugs. It's possible that this changes the institutional threshold for comfort even if nobody consciously says, “We don't approve illegal drugs.”
• FDA may be protecting the precedent, not merely judging MDMA. Approve this evidence package and future sponsors can point to it. The question becomes not only “Are we comfortable with this treatment?” It becomes:  “Are we comfortable with this becoming the evidentiary standard for the next ten psychedelic drugs?”
• The therapy component creates jurisdictional discomfort. FDA could be approving a drug whose effectiveness depends upon therapists, therapeutic methods and clinical environments FDA doesn't regulate. A cautious institution may be reluctant to approve something when part of the intervention lies outside its normal control.
• Institutional path dependence. Regulatory systems are built around conventional pharmaceuticals. A treatment that looks like drug + altered state + therapist + environment + integration doesn't fit into a system designed around drug → biological effect → clinical outcome. Novelty can create friction without anyone opposing the treatment.
• Political/reputational exposure. Imagine the headlines after the first serious adverse event following approval: “FDA Approved Ecstasy Therapy Despite Warnings.” That's a different reputational environment from an adverse event involving an obscure new molecule nobody outside medicine has heard of.
• The sponsor may have weakened the case. Concerns about trial conduct, therapist behavior, data collection and study methodology may have made FDA unwilling to take a regulatory leap on an already-unusual product.

There's evidence against the conspiracy hypothesis

If our hypothetical skeptic says: “FDA doesn't want psychedelics approved,” we can test that hypothesis against subsequent behavior.

It doesn't fit well. FDA designated MDMA-assisted therapy a Breakthrough Therapy years earlier. It worked with MAPS on the Phase 3 protocol. The federal government is facilitating additional psychedelic research rather than trying to shut the field down.

MAPS itself describes expanding federal research activity in 2026 involving MDMA, psilocybin, LSD and other psychedelics.

There's political pressure in the opposite direction. In May 2026, a bipartisan group of 30 members of Congress asked FDA to address issues including predictability of the Special Protocol Assessment process, functional unblinding, advisory-committee expertise and the evolving role of psychotherapy with pharmacological interventions.

I wouldn't, therefore frame the alternative explanation as, the  FDA secretly doesn't want MDMA approved. I'd frame it as a question:

Could FDA's decision reflect not only the scientific deficiencies it identified, but also institutional incentives that make regulators especially cautious about being first?

There's a good test of that hypothesis coming. If another psychedelic drug eventually presents FDA with cleaner trials, better blinding, clearer separation of drug and therapy, stronger safety characterization and fewer study-conduct problems, and the FDA still keeps moving the goalposts, the institutional-bias hypothesis becomes interesting.

If FDA approves it?

The simpler explanation for MDMA's rejection becomes stronger: the MDMA application itself wasn't good enough.

That's why I think the LSD and psilocybin programs are interesting. They aren't merely other potential treatments. They're future tests of competing explanations for what happened to MDMA.
 

Is MDMA Safe?

Safe doesn't mean incapable of causing harm.

Many useful medications can cause harm. MDMA, for example, may increase heart rate, blood pressure, and body temperature.

Jill emphasized that these treatments aren't appropriate for everyone.

She described screening for schizophrenia and other psychotic disorders and said that even family history can matter. That concern isn't unique to Jill. Some MDMA and psychedelic research protocols exclude participants with a personal history of psychotic disorders and even those with a first-degree relative with schizophrenia or another psychotic disorder.

Bipolar disorder can also affect eligibility. Clinical-trial criteria vary, but some protocols have excluded bipolar I disorder, while others exclude both bipolar I and II.

Cardiovascular health is another consideration. MDMA can increase heart rate and blood pressure, and clinical trials have excluded people with certain cardiovascular conditions, including uncontrolled hypertension, some arrhythmias, and other conditions that could make those increases dangerous.

This isn't a checklist someone should use to medically clear themselves for MDMA or another psychedelic treatment. It's the opposite.

Nothing comes without risk. The useful question isn't: Is MDMA safe?

It's, hat are the risks at the doses and under the conditions being proposed for treatment, and for whom might those risks be unacceptable?
 

Jill's Experience Is Evidence - Not Everything

Jill says MDMA-assisted psychotherapy changed her life. I believe that's what she experienced.

Saying Jill got better after MDMA-assisted psychotherapy doesn't establish that MDMA-assisted psychotherapy will make everyone with PTSD better.

Studies may tell us what happened across groups under defined experimental conditions.

Regulatory review can identify weaknesses and risks in the evidence supporting a proposed treatment.

Jill can tell us something neither can.

What was it like to go through it?

Before MDMA Could Be the Answer, Jill Had to Find the Problem

Jill didn't spend her life knowing she had PTSD and searching for an unusual treatment. For years, she thought certain behaviors were simply her personality.

She was anxious. She worked extremely hard. She needed plans A, B, C, D and E. She constantly thought about safety. She didn't understand that she was living in a state of hypervigilance. She thought that was just Jill. Eventually she was diagnosed with PTSD.

That's where her story is useful even if MDMA-assisted psychotherapy is not something you would consider. Before asking, what's the best treatment, there's a more fundamental question:

What are we trying to treat?
 

Problem #1: I Think, This Is Just Who I Am

If I believe I am an anxious person, the problem may appear to be anxiety. If I believe I am an obsessive worker, I might try to improve my work-life balance. If I believe I'm a pessimist, I might try to think more positively.

What if those behaviors are manifestations of something else? I could spend years trying to change the symptoms without understanding what's producing them.

That doesn't mean every personality trait or uncomfortable emotion has a hidden diagnosis. It means that when something significantly interferes with your life, identifying the problem matters before choosing the solution.
 

Problem #2: The Event Doesn't Seem Bad Enough

What counts as trauma? Do you have to experience something obviously catastrophic before your experience qualifies?

Jill redirected the question toward what the person is experiencing rather than ranking the severity of the event. Don't confuse the size of an event with the size of its effect. Two people can experience similar circumstances and respond differently.

Instead of asking, was what happened to me bad enough to count as trauma? Ask, what effect is it continuing to have on me? That question might tell you there's something worth discussing with someone qualified to determine what's happening.
 

Problem #3: Thinking The Treatment Isn't Working

Jill didn't begin with MDMA. 

After her longtime partner died, she went to therapy for grief. She said that work helped.

Later, something different happened. A relatively minor work email triggered an enormous physical response. Intellectually, she understood that the catastrophic outcome she imagined wasn't reasonable. Her body reacted as though it were.

She wasn't sleeping properly. She wasn't eating properly. She was shaking. Her stomach was upset. She had intense physical tension. Eventually she and her therapist began exploring childhood experiences.

After months of conventional therapy without sufficient improvement, she was diagnosed with PTSD. Only then did MDMA-assisted psychotherapy enter the discussion.
 

Problem #4: Judging the Treatment by Its Reputation

When MDMA was proposed, Jill resisted. Essentially, she told them: I don't do drugs. I came into our conversation carrying the same baggage: Ecstasy. Illegal drug. Bad for you. Why would a therapist give someone that?

The opposite bias is possible: MDMA becomes medicine. Now it sounds legitimate. Safe. Therapeutic. Neither label answers the questions we actually need to ask:

What dose?
For what condition?
Under what circumstances?

What are the risks? Who shouldn't receive it?

Who is supervising it?
What happens before and afterward?

What Changed for Jill

During Jill's first MDMA-assisted session, childhood memories surfaced. She described seeing those memories almost as though they were on a television screen.One involved being moved among relatives. As a child, she had interpreted that experience as, I don't deserve a home. Everybody else gets a family. I don't. There's something wrong with me.

Looking at the event as an adult, she reached a different conclusion: Her parents hadn't been capable of taking care of her.Same history. Same memory. Different interpretation.

The treatment didn't change what happened. It changed her ability to examine what happened, without reproducing the same response she'd carried from childhood.

That raises a question that goes well beyond MDMA: How many of our present-day reactions are based on conclusions we formed under circumstances that no longer exist, and are those conclusions still true?
 

MDMA Didn't Do Everything

Jill made clear that the experience wasn't turnkey. She didn't take MDMA, discover a new interpretation of her childhood, and never have to think about it again.

Her treatment didn't happen in an afternoon. Jill described three MDMA-assisted journeys over approximately a year, surrounded by preparation, therapy, and integration. By the summer following her third journey, she described substantial changes.

She was no longer afraid of the universe.
She said she was no longer suicidal.
She wasn't hyper vigilant.

She believed she had a future.
According to Jill, she no longer met the diagnostic criteria for PTSD.

The work didn't end there. She described difficult memories as sometimes needing to be revisited, comparing the process to using flashcards to learn something:
 

New ways of thinking require repetition.

She continued working with what had emerged. She journaled. She walked without headphones and gave herself time to think. She continued reinforcing the changes she believed she'd made.

That reminded me of physical therapy after I broke my hip climbing. Surgery fixed something. It didn't restore my strength and balance. I still had work to do.

A question worth asking about any treatment:

What part of the problem is the intervention supposed to solve, and what part remains mine afterward?

Instead of saying, “MDMA cured Jill's PTSD,” maybe a better description is: MDMA-assisted psychotherapy was part of a year-long process through which Jill says she eventually stopped meeting the diagnostic criteria for PTSD.

Don't Start With MDMA

Don't start with any treatment.

Jill didn't.

Her path looked like this:

Something isn't right.

→ What am I actually experiencing?
→ What's causing it?
→ Do I know that's the problem?
→ Who can help me determine that?
→ What treatments have evidence for helping solve the problem?
→ What are their risks and limitations?
→ Which is appropriate for me?

Then she got to:

What treatment should I choose?

If someone reads Jill's story and concludes, I have anxiety, maybe I need MDMA, they've started at the wrong end. Jill cautioned me against doing that when I tried to broaden the discussion from PTSD to anxiety, depression, and other conditions.
 

What We Have Established, and Have Not

Based on my conversation with Jill and some basic independent thinking and research, I think we can say:

MDMA is the drug associated with Ecstasy and Molly. That doesn't make recreational use and professionally supervised MDMA-assisted psychotherapy equivalent exposures or experiences. That doesn't make recreational use and professionally supervised MDMA-assisted psychotherapy equivalent exposures and experiences.

Dose matters, as do the environment and circumstances in which a drug is used.

MDMA isn't interchangeable with other drugs called psychedelics. Jill's experiences with MDMA and psilocybin were different.

MDMA-assisted psychotherapy isn't just taking MDMA. The treatment that has been studied includes a therapeutic process centered around the experience of taking the drug.

Clinical trials reported substantial improvements in PTSD symptoms. That doesn't mean the evidence is beyond question.

The FDA's refusal to approve the treatment doesn't establish that it doesn't work. It tells us the agency wasn't satisfied that the submitted evidence adequately established effectiveness and safety for approval.

Jill's experience doesn't prove MDMA-assisted psychotherapy will work for someone else. It tells us what happened to Jill.

We don't need to make her story establish something it can't. When information gets compressed, qualifications disappear.If you're personally interested in MDMA-assisted psychotherapy, the first question shouldn't be, should I try MDMA?

It should be:

What am I experiencing, and do I understand what's causing it?

Find someone qualified to help answer that question. 

Sometimes the reason we can't find the right answer isn't that we need another answer.

We need a better question.

Editor's Note: This article is based in part on my podcast interview with Jill Sitnick, author of Rescuing Jill, about her experience with MDMA-assisted psychotherapy. Jill's personal experiences and observations originate from that conversation. The broader discussion also draws on published clinical research and FDA materials. The structure, questions, emphasis, and commentary are my own. Any errors or interpretations should be attributed to me, not to Jill.

This article is intended for general informational purposes and is not medical, psychiatric, or therapeutic advice. Jill's experience should not be interpreted as evidence that MDMA-assisted therapy is appropriate for any particular person. If you have questions about PTSD, trauma, MDMA-assisted psychotherapy, or another mental-health treatment, consult a qualified medical or mental-health professional.

About the Author

Daniel Stih is an aerospace engineer, software engineer, indoor environmental consultant, and author of 12 books. For more than 30 years, he has investigated complex problems spanning engineering, technology, the built environment, and human decision-making. His work explores how evidence, assumptions, and systems shape the conclusions we draw—and whether we're solving the right problem. Learn more about his approach in Why I Think This Way.