The American conversation about psilocybin treats approval as an FDA question, framed around Compass Pathways’ march toward a New Drug Application. That framing is too narrow. The FDA route is only one of the ways the developed world is deciding whether and how patients get psilocybin, and the sharpest divergence is already in place: one wealthy country, Australia, has permitted psychiatrists to prescribe psilocybin since 2023, and it got there without a completed Phase 3 trial. The international picture is not a single regulatory finish line approached at different speeds. It is a set of incompatible models that disagree about what has to be proven, and when.
Model one: prove it, then permit it
The route familiar to a US or European reader is evidence-first. Compass Pathways, a London-based company, has run its COMP360 psilocybin program as a classic registration effort. Its pivotal trials enrolled across North America and Europe, the second of them randomizing 581 participants on two continents, for a combined Phase 3 dataset of more than 800 patients, and the company is targeting a rolling NDA submission in the fourth quarter of 2026. The logic is sequential: complete the pivotal trials, file with the regulator, obtain marketing approval, then allow prescribing of an approved product. Because European sites are embedded in that dataset, a filing with the European Medicines Agency could follow the FDA one. This is the prove-then-permit model, and it makes the Phase 3 the gate through which access has to pass.
Model two: permit it under controls, then observe
Australia did the opposite. Effective July 1, 2023, the Therapeutic Goods Administration reclassified psilocybin, for treatment-resistant depression, and MDMA, for post-traumatic stress disorder, from Schedule 9, prohibited substances, to Schedule 8, controlled medicines, allowing authorized psychiatrists to prescribe them. The crucial detail is what was not required. There are no psilocybin products on the Australian Register of Therapeutic Goods that the TGA has assessed for quality, safety, or efficacy. Access runs to unapproved products, through the Authorised Prescriber Scheme, under controls that include human research ethics committee approval and import permits. Australia, in other words, permitted prescribing before a registration-grade evidence base existed, and built the safeguards around the prescriber rather than the product.
It was contested at the time. Australian researchers who ran the country’s own psilocybin trials publicly objected that the agency had not consulted the clinical experts with direct experience, and noted how thin the local evidence base was when the decision was made. The decision was also a reversal: the TGA had issued an interim decision against rescheduling only months earlier before changing course. This is the permit-then-observe model, and it makes the Phase 3 optional for access.
Why the divergence matters
Same molecule, opposite sequencing. The US and European route makes pivotal evidence the precondition for patients getting the drug. Australia made it possible to prescribe the drug while that evidence was still being gathered elsewhere. The consequence is a fragmented global map in which a developer running a single global Phase 3 program operates alongside a parallel access channel, in a wealthy market, that does not depend on its trials at all. Compass and the deuterated-psilocin developer Cybin, domiciled in Ireland, are both building registration dossiers for the prove-then-permit jurisdictions while Australia already has psilocybin in psychiatrists’ hands through a different door.
For the audience that tracks this commercially, three implications follow. Market entry is not one event but a series of different gates, and in Australia the gate is already open, though it opens onto a controlled prescriber scheme for unapproved product rather than a conventional product market, which makes commercial capture there look nothing like an FDA approval. Real-world prescribing in Australia generates observational safety and outcome data, much as Oregon’s community program does in the US, and that data could either support or complicate the registration narrative being built in other jurisdictions. And Australia’s leapfrog is a precedent that advocates and regulators elsewhere will cite: if it produces good outcomes, it strengthens the case for access-first models in other countries; if it surfaces safety problems, it hands the evidence-first camp its argument.
The limits on both sides
Neither model is as clean as its summary. Australia’s scheme is restrictive in practice, with few authorized prescribers, a heavy approval burden, and high out-of-pocket costs, so prescribing-in-principle has translated into limited access-in-fact. And the two models do not produce the same kind of access even when both function. The US and European route, if Compass succeeds, yields an approved, manufactured, labeled medicine with formal post-marketing controls. Australia’s pathway yields supervised access to an unapproved product. Calling both of them psilocybin access obscures how different the resulting systems are.
The frame
The useful question is not when psilocybin will be approved. It is approved where, and through which logic. The United States and Europe are running the prove-then-permit playbook, with a possible first approval on a roughly 2027 horizon. Australia ran permit-then-observe and has been prescribing since 2023. The coming years will function as a natural experiment in which sequencing produces better outcomes, and the result will shape psychedelic regulation far beyond any single FDA decision. The most accurate description of the present is the one the FDA-centric coverage tends to miss: psilocybin is already a prescribable medicine in one developed country and an investigational drug in the others, and that gap, not the timing of a US filing, is the international story.