Four mental health verticals reached a milestone in the first five months of 2026. A classic psychedelic cleared Phase 3 for the first time. Cannabis moved to Schedule III, partially. A genuinely new antipsychotic mechanism reached the market. Digital mental health kept scaling on a public bellwether. Each of these is a real event. In each case, the coverage describes a larger change than the underlying mechanism delivers, and the gap between the two is where the next two quarters will be decided.
The pattern is worth naming directly, because it repeats across verticals that otherwise share little. The headline event is binary in framing (approved, rescheduled, launched, scaling) and conditional in substance (which dose, which license category, which patient population, which revenue line). Reading the substance, rather than the framing, is the whole job.
Psychedelics: two Phase 3 trials hit, and the hard questions are still ahead
On February 17, Compass Pathways reported that both of its pivotal Phase 3 trials of COMP360, a synthetic psilocybin formulation, met their primary endpoints in treatment-resistant depression. COMP005 compared a single 25 mg dose against placebo and produced a 3.6-point separation on the Montgomery-Asberg Depression Rating Scale at week 6, with a p-value below 0.001. COMP006 compared two 25 mg doses given three weeks apart against a 1 mg comparator and produced a 3.8-point separation, also below 0.001. This is the first time a classic psychedelic has cleared Phase 3. Spravato, Johnson and Johnson’s esketamine, remains the only psychedelic-adjacent product approved, and it is a ketamine derivative rather than a classic serotonergic psychedelic.
That is the milestone. Now the mechanism. A 3.6 to 3.8 point MADRS separation is statistically unambiguous and clinically modest. The MADRS runs 0 to 60, and the field’s existing benchmarks for what counts as a meaningful difference sit higher than what these trials produced on the primary timepoint. Compass has not yet been through FDA review; the company has said it will engage the agency, which means the regulatory questions are in front of the program, not behind it. The two questions this desk has flagged repeatedly in psychedelic trial design, durability past the acute window and functional unblinding (participants and raters correctly guessing assignment because the drug is subjectively obvious), are not resolved by hitting a week-6 endpoint. They are the questions an FDA advisory committee will ask.
The post-Lykos read on the sector was that the regulatory path had closed. Two positive Phase 3 readouts reopen it. They do not de-risk it. An approvable Phase 3 dataset and an approval are different objects, and the distance between them is exactly the methodological territory the agency rejected Lykos over.
Cannabis: Schedule III arrived for some operators and not others
On April 23, the Department of Justice issued an order immediately placing two categories of cannabis into Schedule III: FDA-approved products containing marijuana, and marijuana regulated under a qualifying state medical license. This followed the December 18, 2025 executive order directing the Attorney General to complete rescheduling expeditiously. It is the largest federal shift on cannabis since 1970.
It is also not what “cannabis rescheduled to Schedule III” implies to most readers. The order does not touch recreational cannabis sold under state adult-use licenses, which remains Schedule I. The broader question, whether to reschedule marijuana at large, goes to a DEA administrative hearing that begins June 29 and is scheduled to conclude by July 15.
The commercial consequence runs through Section 280E of the tax code, which disallows ordinary business deductions for sellers of Schedule I and II substances. Under the April order, operators conducting activity under a state medical license can now deduct rent, payroll, and marketing for that activity, moving their effective federal rate from the punitive range toward the standard corporate rate. Recreational activity under adult-use licenses cannot. The result is three operator categories: pure-medical (real relief), dual-license multi-state operators (a genuinely difficult expense-allocation problem between medical and recreational lines), and pure-recreational (no change). Any analysis that treats the April order as uniform relief for the sector is reading the framing and not the order. The June 29 hearing is the binary event that determines whether the bifurcation collapses back into a single rescheduled market.
Pharma psychiatry: a new mechanism arrived, and its biggest market resisted it
Bristol Myers Squibb’s Cobenfy (xanomeline and trospium chloride, formerly KarXT) is the first genuinely new mechanism of action approved for schizophrenia in decades. It is a muscarinic agonist rather than a dopamine antagonist, which is the entire reason it matters: it sidesteps the side-effect profile that has defined antipsychotic non-adherence since the class was created. As a monotherapy, it is approved and on the market.
The mechanism story narrows at the adjunctive trial. In April 2025, the Phase 3 ARISE trial tested Cobenfy added on top of an atypical antipsychotic in patients with inadequately controlled symptoms, and it missed its primary endpoint: a 2.0-point separation on the PANSS total score at week 6, with a p-value of 0.11. Adjunctive use, treating patients already on an antipsychotic who are not responding fully, is where the volume sits in real-world prescribing. The monotherapy approval addresses the cleaner population; the larger commercial opportunity sits in the population where the pivotal adjunctive trial did not reach significance. The safety and tolerability held up, which is why the company signaled it would keep talking to regulators, but the efficacy signal in the high-volume setting is the open question, not the mechanism’s novelty.
Digital: the bellwether kept growing and stopped being a mental health story
Hims and Hers reported Q1 2026 revenue of $608.1 million, up roughly 4 percent year over year, missing the consensus estimate near $617 million, and the stock fell sharply on the print. The company swung to a net loss of about $92 million from net income the prior-year quarter, driven by roughly $33.5 million in restructuring and one-time charges. Subscribers reached 2.58 million, up 9 percent. Full-year guidance was raised to a range of $2.8 to $3.0 billion.
The number that matters for this publication is not the miss. It is the composition. The growth is being driven by branded GLP-1 weight loss products and international expansion, with international revenue up nearly tenfold off a small base. Mental health, the category that anchored the original telehealth-for-psychiatry thesis, is now one specialty inside a platform whose marginal growth comes from weight loss. The 59 percent full-year 2025 growth decelerating to single digits in Q1 2026 is the more important signal than any single quarter’s beat or miss: the standalone “digital mental health as a business” framing is being absorbed into a general consumer-health platform story, and the reimbursement-and-efficacy questions that define digital therapeutics as a regulated category are being answered, commercially, by routing around them.
What connects the four
The verticals do not share a regulator, a mechanism, or a customer. They share a structure. In each, 2026 produced an event that reads as a category-defining win and resolves, on inspection, into something partial: a modest effect size awaiting FDA scrutiny, a rescheduling that covers medical and not recreational, a new mechanism that missed in its highest-volume indication, a growth story migrating away from the category that named it.
Each also has a near-term event that converts the conditional into something more definite. Compass enters FDA engagement on a dataset whose methodology is the live question. The DEA hearing on June 29 decides whether cannabis rescheduling stays bifurcated. Bristol Myers Squibb’s conversations with regulators determine whether Cobenfy’s adjunctive path continues. And the next several quarters determine whether digital mental health is a category or a feature. The coverage will report each of these as another binary headline. The mechanism, in each case, will be narrower and more conditional than the headline. That gap is the thing worth reading for.