The Problem the FDA Couldn't Solve: What MDMA's Rejection Revealed About Approving Therapy

On August 9, 2026, the trade publication Psychedelic Alpha reported that Resilient Pharmaceuticals had resubmitted its application for MDMA-assisted therapy for PTSD to the FDA, without running the additional Phase 3 trial the agency had asked for. MAPS responded the next day.

Worth being precise about what that means. The company itself hasn't announced this. MAPS, which funded the original research, no longer has an active role in the program and was responding to a report rather than confirming its own filing. So the news is real but secondhand, and the details may shift.

What's more interesting is the two-year gap, and what the field learned in it.

What actually happened in 2024

In June 2024, an FDA advisory committee voted 10 to 1 against recommending approval. In August, the agency issued a Complete Response Letter, which is a formal notice that an application can't be approved as submitted, along with what would need to change. Within weeks, the company laid off roughly three-quarters of its staff.

For a year, nobody outside the process could read the FDA's actual reasoning. That changed in September 2025, when the agency released the letter as part of a batch of 89 previously unpublished response letters. For the first time, the specific objections were public.

They fell into a few categories.

Durability. The trials didn't demonstrate that the benefit held past the 18-week end-of-study assessment, which came only eight weeks after the last dose. For a treatment framed as producing lasting change from a small number of sessions, that's a pointed gap.

Safety reporting. The agency raised concerns about unreported adverse events at two sites, and about instructions given to trial sites regarding how to categorize euphoria-like effects. When a regulator loses confidence in how safety data was collected, everything downstream gets harder.

Prior MDMA use. A substantial share of participants had used MDMA before, which raises questions about how well the results generalize.

Bias and blinding. This is the deepest one, and it deserves its own section.

Why blinding breaks

Drug approval rests on the randomized controlled trial, and the RCT rests on blinding. Neither the participant nor the person assessing them should know who got the drug.

With MDMA, this is close to impossible. Participants know. The therapists in the room know. In the trials, most people correctly guessed their assignment. The technical term is functional unblinding, and once it happens, you can no longer cleanly separate the drug's effect from the effect of knowing you received the real thing and expecting to improve.

This isn't a flaw someone introduced through carelessness. It's a property of the treatment. A compound that produces several hours of altered consciousness announces itself.

MAPS's position is that the therapists delivered high-quality care to placebo participants as well, and that the data remained valid despite partial unblinding. Rick Doblin has argued more pointedly that the FDA shifted its standards after having agreed to the trial design years earlier. The FDA's position is in the letter.

Both accounts are on the record. They don't reconcile, and I don't think an outside observer can settle it. What's clear is that the disagreement was not really about whether MDMA does something. It was about whether the evidence was assembled in a way the approval system could accept.

The part that concerns therapists most

Here's the structural problem underneath everything else.

The FDA approves drugs. It does not approve psychotherapy, and it has no established framework for doing so. But MDMA-assisted therapy was never a pill. It was a compound delivered inside a specific structure: preparation sessions, two trained therapists present through an eight-hour dosing session, integration sessions afterward. The therapy wasn't packaging around the drug. It was constitutive of the treatment.

That created a question the system wasn't built to answer. If the treatment works, what worked? And if the therapy is a necessary component, who certifies the therapists, who monitors quality, and what happens when the treatment leaves a controlled trial and enters ordinary practice?

The advisory committee circled this repeatedly. The agency has since put it in writing.

What changed in 2026

The regulatory picture now looks quite different from 2024.

In April 2026, an executive order made accelerating psychedelic treatments for serious mental illness a stated federal priority, directing the FDA toward faster review and expanding research funding.

In July 2026, the FDA published final guidance titled Psychedelic Drugs: Considerations for Clinical Investigations, finalizing a draft that had been sitting since June 2023. The guidance directly addresses the questions the MDMA rejection raised: trial design, safety characterization, therapist-delivered interventions, and long-term follow-up. It suggests approaches to the blinding problem, including active comparators such as lower doses or other psychoactive compounds that produce some subjective effects, and the use of central raters who are blinded to assignment.

Legal analysts reading the final version have noted it both opens a path and sets a high evidentiary bar. It also shifts some of the safety burden into the post-approval period, pointing sponsors toward risk management programs rather than requiring every question to be settled before approval.

Alongside it, the federal health agencies asked for public input on how psychedelic treatments would actually be delivered in outpatient settings if approved, and the FDA scheduled a public hearing on the therapeutic use of these compounds for September 14, 2026.

There's a political dimension here worth naming without overinterpreting. The acceleration is coming from the executive branch, and the company at the center of the MDMA application was recapitalized in 2025 by investors with political connections. Faster pathways can mean earlier access for people who need it. They can also mean pressure on an agency that has just been criticized for the opposite failure. Both can be true, and it's early to know which dominates.

What I take from it

A few things I hold onto as someone who works clinically in this territory.

The rejection was not a verdict on whether MDMA helps people with PTSD. It was a verdict on whether one application, assembled a particular way, cleared the bar for approval. Those are different questions, and conflating them has caused a lot of unnecessary despair and a fair amount of unnecessary triumphalism.

The hardest problems in this field are methodological, not pharmacological. We are reasonably good at studying molecules. We are not good at studying molecules that are inseparable from a relationship, a setting, and hours of skilled human accompaniment. That's not a psychedelics problem. It's a problem psychedelics happen to make impossible to ignore.

The therapy question is still unanswered. The guidance acknowledges therapist-delivered interventions as a factor to account for. It doesn't establish who trains, credentials, or oversees the people delivering them. That will be decided somewhere, by someone, and clinicians should be paying attention to where.

And the timeline is genuinely uncertain. An application has reportedly been resubmitted. That starts a review, not a countdown. Anyone telling you they know how this resolves is guessing.

For those of us working with what's already legal, the practical takeaway is smaller and steadier. The care model that surrounds the medicine, the preparation, the presence, the integration, is the part that transfers across compounds and across regulatory outcomes. That work doesn't wait on the FDA.