The race to approve psilocybin

A nonprofit and a public company are each close to asking FDA to approve the same molecule for depression. Who gets there first matters less than you'd think. The first approval sets a precedent. What comes after it, including the second approval and the competition between them, may decide what psychedelic treatment actually looks like in everyday psychiatry.

By Reid Robison, MD. 3,152 words, with where each program stands, the evidence side by side, an FDA review clock and the treatment day compared.

Researched and reviewed by Reid Robison, MD × OpenEvidence × Claude How this post was made

1Where things stand

Psilocybin isn't new. What would be new is an FDA-approved psilocybin product with a label, a standard dose and a regulated way of giving it. Two organizations are closest.

Usona Institute, a Wisconsin nonprofit, is developing psilocybin for major depressive disorder. Its phase 3 trial, uAspire, randomized adults to a single 25 mg dose, a 5 mg dose or an inactive placebo, with outcomes through day 43. Enrollment is complete.⁷⁸ As of today, the NDA report is Psychedelic Alpha's, not an announcement from Usona or FDA.¹

Compass Pathways, a Nasdaq-listed company, is developing COMP360, its own formulation of psilocybin, for treatment-resistant depression. FDA agreed to a rolling review in April, some modules are in, and Compass expects to submit the rest this quarter. It has guided to a launch in the first half of 2027, if FDA approves and DEA reschedules.² FDA site inspections began in September.¹

So the race has three finish lines, not one: the first complete application, the first approval and the first launch. They may not have the same winner.

2Two programs, one molecule

Same active drug, different bets.

The condition may be the biggest difference. Treatment-resistant depression is a smaller, harder-to-treat slice of major depression. If both are approved, the labels may point at different patients, which affects who gets referred, what insurers cover and where psilocybin sits in the line of treatments.

3The evidence, side by side

If you read my fair test post, the first question is "compared with what?" Here are the published effects, each as the difference from that trial's control on the MADRS scale.

Four things keep this from being a race chart.

• Different people. Usona's phase 2 enrolled adults with major depression; Compass enrolled people whose depression hadn't responded to at least two treatments. Harder-to-treat groups usually show smaller effects.
• Different controls, and they cut both ways. Usona's phase 2 used niacin, which flushes the skin but has no psychedelic effect; COMP006 used 1 mg of psilocybin; COMP005 used an inert placebo. A control that feels like something can shrink the measured difference. The bigger worry usually runs the other way: when people can tell they got the control, the control group may improve less than a typical placebo group, which makes the drug look better. In Compass's phase 2b, the 1 mg group improved 5.4 points by week 3, less than placebo groups often do in antidepressant trials.⁵ And COMP005, with an inert placebo, still showed the smallest effect of the three.
• Different timepoints and sizes. Usona's 12.3 points was measured at day 43 in 104 people, so its 95% confidence interval runs from 7.2 to 17.5. Compass's phase 2b number was at week 3 (6.6; 2.9 to 10.2, 233 people), and its phase 3 numbers at week 6 (3.6; 1.5 to 5.7, and 3.8; 1.8 to 5.8, in 258 and 581 people).⁴⁵⁶
• Different phases. Phase 2 effects often shrink in phase 3. Compass's own phase 2b showed 6.6 points at week 3; its phase 3 trials showed 3.6 and 3.8 at week 6.⁵⁶ I walked through that pattern in the last post. It's a general pattern, not a forecast for any trial, and Usona hasn't released its phase 3 results.

Not every trial has been positive. EPISODE, an independent German trial in 144 people with treatment-resistant depression, missed its primary endpoint: response at week 6 wasn't significantly higher on 25 mg than on an active placebo (nicotinamide), though it was at week 1, and average depression scores at week 6 were 4.6 points better on 25 mg.⁹ That's a useful reminder of how unsettled the size of the effect still is.

Compass also reported that 39 percent (COMP006) and 25 percent (COMP005) of people on 25 mg had a "clinically meaningful response," which it defined as at least a 25 percent drop in MADRS.⁶ That's a useful number, and it's worth knowing it's a lower bar than the usual definition of response in depression trials, a 50 percent drop. So it can't be compared directly with response rates from other trials. In COMP006, people who responded kept their improvement on average through 26 weeks, in an analysis Compass describes as post hoc, and 58 percent of the 25 mg group were retreated after week 9.¹⁰ In the phase 2b, the difference in sustained response at week 12 wasn't statistically significant.⁵

Safety, briefly. Usona's phase 2 reported no serious treatment-emergent adverse events, though psilocybin caused more side effects overall than niacin, including more rated severe.⁴ In Compass's phase 2b, suicidal ideation, suicidal behavior or self-injury occurred in all three dose groups and was more common on 25 mg and 10 mg than on 1 mg; the authors called for careful monitoring.⁵ In COMP006 through 26 weeks, serious adverse events occurred in 5.7 percent on 25 mg, 3.5 percent on 10 mg and 6.3 percent on 1 mg. There was one suspected suicide in the 25 mg arm, judged unrelated by the investigator, and suicidal ideation as a side effect was less common on 25 mg (5.1 percent) than on 1 mg (9.8 percent).¹⁰ Most side effects, mainly nausea, headache and anxiety, happened on dosing days and resolved within a day.⁶ Psilocybin also raises blood pressure and heart rate for several hours; in EPISODE, one of the two serious reactions was a marked rise in blood pressure.⁹

4How fast could FDA move?

Very fast, on paper. A standard review takes about 10 months and a priority review about 6, plus 60 days for FDA to file a new molecule.¹¹ The national priority voucher aims for one to two months after a complete application.³ Then DEA has to move the approved product out of Schedule I. Under a 2015 law, DEA has 90 days to issue an interim rule once it has both the approval and HHS's scheduling recommendation, and Compass says states covering about 90 percent of US patients plan to follow within 30 days.²

A voucher sets a goal; it doesn't lower the bar. FDA still has to agree the evidence is enough, finish inspections, settle the label and approve a REMS, the safety program that governs where and how a drug like this can be given. A first-of-its-kind psychiatric drug could raise questions that take longer.¹²

5The treatment day

An approved psilocybin product won't be picked up at a pharmacy. It will be given in a clinic, under supervision, for most of a day. How much support goes with it is one of the most important open questions, and the two programs have leaned different ways.

That difference raises a question FDA will have to answer in the label and REMS: is psilocybin a drug that needs monitoring, or a treatment that depends on a psychological-support protocol? FDA's own guidance calls the contribution of that support an open scientific question.¹² I wrote about how much of the change might be the molecule versus the people in the room in another post.

The answer will decide who can offer it. A clinic needs a room for most of a day, one or two trained staff, screening (including for heart disease and blood pressure), blood pressure checks during the session, emergency procedures and secure storage for a controlled substance. A lighter model would scale more easily. A heavier one may better match what was tested. I looked at what that day might look like in an existing Spravato clinic in this post.¹⁴

6A nonprofit and a public company

This is the part of the story I find most interesting, and it doesn't have a villain.

Usona is funded by philanthropy, with a stated mission around research and access. That gives it room to make choices a company might not. But being a nonprofit doesn't make a drug cheap. Manufacturing, safety monitoring, a REMS, distribution and insurance contracting all cost money, and I couldn't find a public price or distribution plan from Usona.

Compass has built the things a launch needs: capital (about $433 million in cash in June, which it says lasts into 2028) and partnerships with Radial and Osmind meant to support treatment sites.² It has described a per-dose price, billed through the clinic, benchmarked to Spravato, but hasn't announced a number.¹⁴ Yes, a public company has to earn a return. And that's also how many hard treatments get built out at scale.

If both are approved, patients and clinicians may get something unusual: a choice between two versions of the same molecule, developed under two different models, competing on access, price and support. That competition may matter more than who was first.

7Questions I'm watching

1. How much is expectancy? Randomization makes the groups alike at the start, but it can't stop expectations once people can tell which group they're in, and most people can tell when they've had a full dose. Remote, independent raters help. Few psilocybin trials have reported whether the blind held.The blinding chart in the last post shows how big that gap can be.
2. How long does it last? And when it fades, does another dose work as well? Durability claims should say whether they cover everyone randomized or only people who responded.
3. How much support is needed? We don't know how much preparation and integration are required to reproduce trial results outside trials.
4. Better than what we have? These trials compared psilocybin with controls, not with esketamine, ECT, TMS or a good next antidepressant. A positive trial shows it works, not that it works better. The one head-to-head with a standard antidepressant, a phase 2 trial against escitalopram, found no significant difference on its primary outcome, and a later analysis suggested psilocybin's apparent advantage on other measures depended partly on what patients expected from escitalopram.¹⁵¹⁶See Ketamine and Spravato, explained for one of the comparisons people will make.
5. Will people have to stop their antidepressants? The trials generally required tapering off them first. If the label does too, that's a real barrier for people who are partly helped by their current medication. A small open-label Compass study gave psilocybin alongside an SSRI.¹⁷
6. Who does it work for outside trials? Trials exclude many people seen in real clinics: bipolar disorder, psychosis risk, active substance use, significant heart disease or uncontrolled blood pressure, and other medical conditions. Post-approval studies will matter.

8What to watch next

• November 3. Compass's quarterly update, which may say when its application is complete.¹ I'll update this post if it changes the picture.
• Late 2026. Whether Usona confirms its filing and whether FDA accepts it.
• Late 2026 to early 2027. The first FDA decision, if the fastest timelines hold, along with the label and REMS.¹
• First half of 2027. Compass's target for launch, if approved and rescheduled.²
• Anytime. Usona's phase 3 results, in a release, a paper or FDA's review documents.

I've spent much of my career inside trials like these, and I'm hopeful. The approval would be historic. But the measure that matters will come later and more quietly: whether a person with depression in an ordinary town can get this treatment, safely, from someone trained to give it, at a price their insurance will cover.

How this post was made

I use AI tools in building this site, and I'd rather be open about how. For this post, I set the scope and the framing and drew on my own experience running trials. Claude helped research the public record, draft the text and build the charts. OpenEvidence, an evidence search tool for clinicians that cites the medical literature, reviewed the draft against the published trials; its critique added the negative EPISODE trial, the escitalopram comparison, a fuller safety picture and a sharper point about blinding. Then I reviewed every number against the sources below. Nothing here comes from non-public trial information. Any errors are mine.

References