The number reported almost everywhere was one: one more country has approved MDMA as a medicine. The number that describes what actually happened is two, and it refers to psychiatrists.
On 4 September New Zealand’s regulator permitted two named prescribers to supply pharmaceutical grade MDMA to patients with severe post traumatic stress disorder, alongside psychotherapy. That is the decision. It is not an approval of a drug for general prescription, and the distance between those two descriptions is the subject here.
What was actually authorized
Regulators have more than one setting. A medicine can be registered for general use, in which case any appropriately qualified doctor may prescribe it within the approved indication. Or access can be granted through a narrower route, in which named practitioners are permitted to supply an unregistered product to particular patients under conditions the regulator specifies.
The New Zealand decision is the second kind. Two psychiatrists, a defined indication of severe PTSD, pharmaceutical grade material rather than anything else, and a requirement that the drug be given within a course of psychotherapy rather than dispensed.
That last condition is not an administrative detail. In the trials that produced the evidence base, MDMA was never the treatment. The treatment was a course of psychotherapy, typically several preparatory sessions, two or three long dosing sessions with two therapists present throughout, and a series of integration sessions afterwards. The drug was an adjunct intended to make the therapy possible for people who cannot otherwise tolerate revisiting a trauma.
Reporting it as approving MDMA describes the smaller half of what is being authorized and omits the part that costs the most.
The date worth looking back at
New Zealand is the second country to do this. Australia was first, in 2023, when its regulator reclassified MDMA and psilocybin so that authorized psychiatrists could prescribe them for specified conditions.
That was reported at the time as a breakthrough, and three years of hindsight make it a more useful guide to what happens next than any trial result.
What followed in Australia was not a wave of treatment. The number of authorized prescribers stayed small. The cost to a patient of a full course, involving many hours of two therapists’ time, ran to tens of thousands of dollars and was not generally covered. Practitioners had to be trained in a protocol that few had been trained in, and there was no established workforce to draw on.
The constraint, in other words, turned out not to be the legal status of the drug. It was the therapy: the hours, the people qualified to deliver them, and who pays.
Anyone reading this month’s decision as the moment New Zealand gained access to a treatment should hold that alongside it. A regulatory permission removes one obstacle. It does not create clinicians.
What the evidence does and does not show
The trial results in this area are genuinely striking and they need to be stated with their limits attached, because they are routinely quoted without them.
As with other results that arrive ahead of the systems meant to carry them, randomized trials of MDMA assisted therapy for PTSD have reported substantial reductions in symptoms against placebo with therapy, in participants with severe, chronic PTSD who had often not responded to other treatments. That is a meaningful result in a condition where existing options frequently fail.
Three cautions travel with it. The trials were relatively small, as trials in this area are. Blinding is close to impossible, because participants can tell whether they have received MDMA, and unblinding inflates measured effects in both directions. And participants in trials of this kind are selected, prepared and supported to a degree that ordinary clinical practice does not reproduce.
None of that makes the finding unreal. It means the effect size observed under trial conditions is the ceiling rather than the expectation, and a regulator granting narrow access to named practitioners is behaving consistently with that reading rather than against it.
Why two is the right number for now
It would be easy to read the restriction as timidity. The better reading is that it is the design.
A limited authorization creates a small number of supervised courses of treatment, delivered by identifiable people, whose outcomes can be recorded and examined. That is close to the only way to find out what a therapy does outside a trial, because the alternative, broad approval followed by observational study, produces data confounded by everything that varies in ordinary practice.
It also fails safely. If the results in a handful of patients are poor, or if adverse effects appear that the trials were too small to show, the exposure is a handful of patients. Broad access forecloses that.
The cost of the design is equally clear and falls on people who are ill now. Severe treatment resistant PTSD is a condition with a substantial mortality, and a person who meets the criteria and cannot reach one of two prescribers is being asked to wait on a process that exists to protect people including them. That is a defensible trade and it is a trade.
There is one more asymmetry worth naming. A drug can be manufactured to a standard and shipped, and a country that decides to import more of it can do so in weeks. A therapist trained in a specific protocol cannot be imported, manufactured or hurried, and two of them are needed in the room at once for sessions that run for hours. Every constraint in this field that gets described as regulatory turns out, on inspection, to be a constraint about hours of trained human attention, which is the least elastic input in medicine.
What would move it
The question worth following over the next two years is not whether more countries follow. Several probably will, and each announcement will be reported the same way.
It is whether any health system decides to fund the therapy hours.
Everything about the scale of this rests there. A course requiring dozens of hours of two trained therapists is expensive in a way that a prescription is not, and no regulator can authorize a budget. Until a public system or an insurer decides that the hours are worth paying for, and until enough clinicians are trained to deliver them, an authorization of this kind produces a small number of treated patients and a large number of eligible ones.
The figure to watch is therefore not the count of countries. It is the count of completed courses, which in Australia after three years is small enough that it is still being reported as individual cases. New Zealand starts with two prescribers. The interesting number will be how many there are in 2029, and whether anybody other than the patient is paying.
