Clival chordoma research
Home › Diet, supplements and claims

Psilocybin as a painkiller

Written 14 September 2026, after a friend asked for the research on psilocybin for pain.

Short version. Psilocybin is not a painkiller. There is no randomised trial anywhere, in any illness, showing it lowers pain intensity. What it does have is four good randomised trials in cancer patients showing large and lasting reductions in depression and anxiety, and a small, mostly negative set of trials in migraine and cluster headache.

There is also a reason specific to the patient to leave it alone, and it is not the one people expect. It is not safety and it is not the law. It is that the way psilocybin is proposed to work on pain is the same thing she has just told us she dislikes about morphine.

1. What the evidence actually is

The best direct tests of psychedelic analgesia used LSD, not psilocybin, in healthy volunteers, with their hands in cold water.

One was positive. Twenty-four people, single doses of 5, 10 and 20 micrograms of LSD against placebo. At 20 micrograms they tolerated the cold longer and rated it less painful and less unpleasant. Blood pressure went up, within normal limits.

One was null. Fifteen micrograms, four doses, randomised and placebo-controlled. No analgesic effect on any measure. A subgroup analysis found something after the first dose only, which is the kind of finding that generates the next study rather than settling this one.

That is the whole of the direct experimental evidence. Cold water in a healthy volunteer is a long way from a 3.8 cm mass on the clivus.

The clinical signal is in headache, and it is weaker than its reputation.

All of it comes from one group at Yale. Three randomised trials:

That last trial is the most honest one in the field and it is the newest. One group, forty-two patients in total, one nominally positive result and two that were not.

And it would not transfer to the patient even if it were solid. Migraine and cluster headache are primary disorders of the trigeminal system. The patient's pain comes from a mass pressing on and growing into nerves. We already checked her eye pain against the cluster headache pattern on Saturday night and it failed on all three counts: the pain was constant, no watering eye, no blocked nostril.

2. Where psilocybin genuinely does work, and why it is the problem

Four randomised trials in people with cancer, including two good ones from Johns Hopkins and NYU, show large reductions in depression and anxiety that were still there four and a half years later. These are real findings by proper researchers and they are why anyone takes the drug seriously.

Every one of them measured psychological distress. None measured pain.

And when you read how psilocybin is proposed to help pain, the mechanisms listed are pain acceptance, cognitive flexibility, and a changed relationship to the pain signal. In other words, it is proposed to work on how much the pain bothers you, not on how much pain there is.

On Saturday the patient said she wants off the morphine, and why: "It doesn't actually work as a painkiller. It just makes me care less about the pain."

That is the same dimension. If psilocybin worked in her exactly as its advocates describe, it would do a thorough version of the thing she has just objected to.

The fairest counter-argument, because there is one. There is a difference between being blunted every hour of every day by a drug that needs increasing, and a single supervised session that shifts something durably with no daily dose. The Yale trials found, oddly, that the benefit did not correlate with how intense the trip was, which hints the two can be separated. If that holds, the affective benefit might be available without living inside it.

It is a fair point. It rests on twenty-four patients in a different disease, and it still leaves the sensory pain, which is the part she actually needs touched, completely unaddressed.

3. Safety, honestly

The general picture is better than the drug's reputation. A 2024 review of 214 studies and 3,504 participants found no serious adverse events at all in healthy volunteers, and no deaths in modern trials.

But in people with a pre-existing neurological or psychiatric condition, serious events occurred in about 4 per cent, and the list includes convulsions. The common effects across every review are raised blood pressure, headache, nausea and vomiting.

Now put the patient against that list:

One piece of good news: her current medicines are not a serotonin interaction risk. Morphine, gabapentin, diclofenac, omeprazole, paracetamol and melatonin are all fine on that front. If amitriptyline or duloxetine gets added later for the nerve pain, that changes.

4. The law, which settles it practically

Psilocybin is Class A and, more importantly, Schedule 1. In UK law that means no recognised medical use. It cannot be prescribed by anyone, for anyone, under any circumstances. There is no compassionate access route. The only lawful way to receive it is inside a licensed trial.

I checked the trial registry today. There are three psilocybin trials recruiting in the UK: one in self-harm, one in depression, one in healthy volunteers. None in pain. The cancer pain trials are both in the United States and the neuropathic pain trials are both in Canada, and none would accept a patient with an unbiopsied skull base tumour and possible seizures.

Which leaves buying it illegally and taking it unsupervised, at an unknown dose, with nobody in the room trained to handle it. That is where the real harm in this area happens and it is worth saying out loud rather than dancing around.

5. What I would rather we spent the week on

This is the actual reason to say no. There are four conventional options for the patient's pain that are evidenced, legal, and still not obtained:

  1. Gabapentin is on 900 mg a day. The licensed maximum is 3,600. No titration schedule has ever been written for her. This remains the single biggest untapped option in the house.
  2. Dexamethasone, asked for and not yet answered.
  3. A sphenopalatine ganglion block, which fits her anatomy better than anything else on the specialist list.
  4. Switching from morphine to oxycodone, converted from her true daily total and not from the background dose alone.

Every one of those has better evidence in nerve pain from a tumour than everything in this note put together, and every one is obtainable from a prescriber this week.

6. The answer in three lines

Where the numbers come from

Full working, with the evidence levels and the parts I am less sure about, is in research thread 23.

Brief for the family · source file em-docs/em-psilocybin-pain.md · last edited 14 September 2026