What a psychotherapist's own MDMA session, and the trial data underneath it, suggest about the real active ingredient in psychedelic therapy.

Sebastian Salicru had been a psychotherapist for over thirty years, and had spent roughly forty of them in and out of his own therapy. So when he took part in a supervised MDMA-assisted group session as part of his training, he wasn't expecting to be ambushed by his own past.

Then a memory surfaced that he was certain he'd long since resolved: himself at four years old, handed over to his grandparents, the warmth of his parents' embrace replaced by an absence he couldn't name. He'd sat with that memory for decades. Explored it, grieved it, filed it under done.

Under the MDMA, he realised he'd been wrong. The child was still there, still waiting — not for more analysis, but for something his years of insight had never delivered. What resolved it, by his account, wasn't the drug. It was the thing the drug made possible: turning toward that younger self with warmth instead of scrutiny, and holding him without needing to fix him.

Salicru published that account this year as a first-person case study. It's a single, subjective story, and he's careful to frame it that way. But it lands on a question the field is only beginning to take seriously — and, unusually, there's now trial data that speaks to it.

The drug opens the door. Something else walks through it.

The intuitive story about MDMA-assisted therapy is pharmacological: the drug quiets the fear response, trust rises, and trauma becomes approachable. True enough. But it leaves the mechanism vague. What, specifically, is doing the healing?

One answer keeps surfacing in the data: self-compassion — and, just as importantly, the retreat of its opposite, the harsh self-judgment trauma leaves behind.

A 2024 analysis of the Phase 3 MDMA trial (van der Kolk and colleagues) found that MDMA-assisted therapy improved self-compassion far more than therapy with placebo — and that this improvement occurred independently of the drop in PTSD symptoms. That independence is the first real clue. It suggests self-compassion isn't just a pleasant side-effect of feeling better; it may be its own lever.

A 2025 follow-up on the same trial (Agin-Liebes, Zeifman and Mitchell) went further and tested it directly. Using the same participants, it asked whether changes in self-compassion statistically mediated the therapy's effect on PTSD and depression. They did — fully. In mediation terms, once you accounted for the shift in self-compassion, the direct effect of the treatment on symptoms was no longer statistically significant. Self-compassion wasn't riding alongside the improvement. It was, on this analysis, the route the improvement travelled.

Two caveats keep this honest. First, it's a secondary, exploratory analysis of a single trial, with self-compassion and symptoms measured at the same time point — so it can't fully prove that the compassion shift came first. Second, and more interesting: the heavier lifting was done by the reduction in uncompassionate responding — self-judgment, isolation, over-identification — rather than the rise in warmth. The largest effects were for self-judgment falling away. So the mechanism may be less "learn to love yourself" and more "stop attacking yourself." Which is almost exactly what Salicru describes: not manufacturing affection for his younger self, but letting the criticism go quiet enough for something kinder to have room.

The medicine opens the door. Self-compassion is what walks through it — and what makes the healing hold.

The uncomfortable part

If the active ingredient is a shift the patient has to undergo — a felt reorientation toward their own pain — then a fair question follows. Can a therapist reliably guide someone through that terrain if they've never crossed it themselves?

Salicru clearly thinks his own session made him a better clinician. He's not alone: in one survey, 88% of psychedelic therapists reported having used a psychedelic at least once, most often for exactly this kind of personal and professional insight. A recent study of Australian psychologists found near-unanimous agreement that first-hand experience would help — one put it as "you can only take your clients as far as you've been yourself." The comparison they reach for is telling: not psychiatry, where no one expects a doctor to have taken the antidepressant they prescribe, but psychoanalysis and mindfulness, where undergoing the process yourself has long been treated as part of the training.

Here's the catch, and it's worth stating plainly because it cuts against the romantic version of this argument. There is no evidence yet that a therapist's own psychedelic experience actually improves patient outcomes. As a 2024 review by Daniel Villiger sets out, no study has ever tested it. What we have is a wall of self-report — therapists saying it helped — and the same pattern holds for therapists who've had the disorder they treat, or done their own therapy: they say it matters, patients often say they prefer it, and yet the hard outcome data stubbornly refuses to confirm it. Belief, preference, and plausibility all point one way. Proof isn't in yet.

So the honest claim isn't "a therapist who hasn't done the work is missing something." It's sharper and less comfortable than that: the whole field increasingly believes experiential knowledge matters, is building training around that belief, and doesn't yet have the evidence to justify it. That's a strange place for a medical field to stand.

Why this should worry you

Now widen the lens, because this is where it stops being an academic debate.

Think about dog training — a field where "force-free" and "e-collar" camps sit at genuinely opposite poles, where someone can complete a weekend course, print a certificate, and legitimately call themselves a professional. It's barely regulated. To an untrained eye, every trainer looks equally qualified. Someone who knows the field can spot the difference in minutes; almost no one else can.

Psychedelic therapy is walking into exactly that landscape, with far higher stakes. The work depends heavily on the person in the room — on their capacity to hold someone through a frightening, defenceless, wide-open state. The literature is candid that a poorly attuned guide can make the experience worse, not just less good. And demand for trained therapists already outstrips supply, which is precisely the condition under which thin credentials rush in to fill the gap: a ten-hour online certificate standing in for lived depth and years of clinical judgment.

That's the real risk in the "does the therapist need experience?" question. It isn't really about whether every therapist must take the drug. It's that we're professionalising a high-stakes, low-barrier field faster than we can define what competence even means — and asking vulnerable people to tell the difference between a Salicru and a weekend-certificate operator, at the exact moment they're least equipped to.

The medicine opens the door. Whether the person sitting beside you can actually help you walk through it — and how you'd even know — is the question the field hasn't reckoned with yet.

Sources: van der Kolk et al., "Effects of MDMA-assisted therapy for PTSD on self-experience," PLOS One (2024); Agin-Liebes, Zeifman & Mitchell, "Self-compassion mediates treatment effects in MDMA-assisted therapy for posttraumatic stress disorder," European Journal of Psychotraumatology (2025); Salicru, "A Healing Experience with MDMA: A Psychotherapist's Mini-Autoethnographic Case Study," Journal of Transpersonal Research (2026); Negrine et al., "'You can only take your clients as far as you've been yourself,'" Drugs: Education, Prevention and Policy (2026); Villiger, "Personal psychedelic experience of psychedelic therapists during training," International Review of Psychiatry (2024); Nielson & Guss, Journal of Psychedelic Studies (2018).