The hidden therapist

Music is the one thing we deliberately put inside a psychedelic session. How a sound wave becomes a feeling, what psychedelics change along the way, and why the playlist may matter more than we admit.

By Reid Robison, MD. About 20 minutes, with seven things to try. Headphones help.

Walk into almost any modern psilocybin or LSD study on dosing day and you'll find the same scene. Someone is lying on a couch with an eye mask and headphones on. One or two therapists sit nearby and mostly stay quiet. For the next six hours or so, the main thing reaching that person's brain from the outside world is music.

We spend years on everything else: the molecule, the dose, who gets screened in and out, the therapy before and after. The playlist is often chosen by whoever on the team has the best taste. Researchers at Imperial College London gave the music a name that stuck: the hidden therapist.¹

I first put this material together as a companion to an episode of Psychedelic Therapy Frontiers, the podcast I host with Steve Thayer. This version follows one sound wave from the air, through the ear and into the brain's emotional machinery, then looks at what changes when a psychedelic is on board, and at what the evidence does and doesn't support.

Part 1Sound becomes electricity

Music starts as physics. A string, a reed or a voice vibrates and pushes the air into waves of pressure. Those waves move the eardrum, which rocks three tiny bones, which push on the fluid inside the cochlea, a spiral about the size of a pea. Along its length sit rows of hair cells. When a hair cell bends, channels in its membrane open, and a wave of pressure becomes a change in voltage. From that moment on, music is electricity.

The scale of that doorway is surprising. Each cochlea has about 3,500 inner hair cells, which carry almost all of what we hear. Each retina has well over 100 million photoreceptors.⁸ Every chill, every sad song, every piece of music that has ever moved you came in through a few thousand cells. Out of that narrow input the brain pulls pitch, rhythm, tempo, loudness, contour and timbre, and builds them back up into melody, harmony and meaning.

Part 2The climb through the brain

The signal doesn't reach awareness all at once. It climbs, and at each level the brain asks the sound a different question. The brainstem has already responded within about 10 milliseconds, long before you know you've heard anything, and the first cortical responses follow within a few tens of milliseconds.⁸

Two of those stations deserve a second look. The auditory cortex is wired tightly to the motor system, which is part of why it's hard to hear a groove without wanting to move. We seem to understand music partly by rehearsing it silently in the parts of the brain that would play it.⁹ And the emotional regions, the amygdala among them, respond to qualities like dissonance and sudden change quickly, before we can say what we're feeling.¹⁰

Life seems to go on without effort when I am filled with music.George Eliot, The Mill on the Floss, 1860

Part 3The chemistry of a chill

By the time a song moves you, it has borrowed chemistry that evolution built for food, safety and belonging. The best-studied piece is dopamine. In a PET study, people listened to music that reliably gave them chills. Dopamine was released in the caudate during the build-up, in the seconds before the best moment, and in the nucleus accumbens when the moment arrived.² Anticipation and arrival are two separate rewards.

That's also why a delayed resolution can feel so good, and why the brain keeps wanting new music once it has learned the old patterns too well. Expectation and surprise are the raw material. A 2025 analysis of a depression trial found that psilocybin therapy and escitalopram changed the brain's response to musical surprises in different ways: escitalopram tended to flatten the emotional difference between expected and unexpected moments, while psilocybin therapy preserved it.¹¹

Dopamine isn't the only chemistry involved. Here's where the evidence stands on the others:

Endogenous opioids

Singing, drumming and dancing together raise pain thresholds, used as a stand-in for endorphin release. Performing raised them more than passive listening.¹²

Oxytocin and bonding

Group singing speeds up bonding between strangers.¹³ Studies of oxytocin itself are small and mixed.¹⁴

Cortisol and stress

The most clinically solid finding. In a meta-analysis of randomized trials, music around surgery reduced pain, anxiety and the amount of pain medicine people needed.¹⁵

Serotonin

The least understood link, and the most relevant here. Serotonin neurons in the brainstem project throughout the auditory pathway and change how sounds are processed.¹⁶ That is the doorway psychedelics walk through.

Hold onto the stress finding. Anxiety going into a session is one of the more consistent predictors of a difficult psychedelic experience.¹⁷ The calm that music can bring isn't decoration. It's part of the clinical work.

Part 4When the medicine enters

Classic psychedelics such as LSD and psilocybin work mainly by activating serotonin 2A (5-HT2A) receptors, which are dense in the cortex. Because serotonin helps tune hearing from the brainstem up, activating these receptors doesn't only change how you feel. It changes how you hear.

• Sound processing shifts. In an MEG study, LSD changed how the brain adapted to repeated tones and responded to unexpected ones.¹⁸
• Music feels more emotional. Under LSD, volunteers rated music as evoking more wonder, transcendence, power and tenderness than on placebo.³
• Music feels more meaningful. LSD increased the personal meaning people attached to songs that had previously meant nothing to them. Blocking 5-HT2A receptors with ketanserin prevented it.⁴
• The brain follows the music differently. LSD changed how brain activity tracked the shifting tonal structure of music, also through 5-HT2A receptors.¹⁹
• Hard feelings may land more softly. Psilocybin reduced the amygdala's response to negative images, and the drop tracked better mood.²⁰

Psychedelics amplify whatever is already there. The music helps decide which way that goes.

Part 5Where music and medicine meet

Music recruits a spread-out network: hearing, emotion, memory, the sense of self, the sense of the body. Psychedelics act on many of the same regions, and that overlap is the best explanation we have for why the two work together.²¹ The map below places the main regions on a side view of the brain. Several sit deep inside or on the inner surface, so treat the positions as approximate.

Four ways the overlap may matter in therapy:

Imagery and memory

Under LSD, music increased the flow of information from the parahippocampus, a gateway to memory, toward visual cortex, and that change tracked how vivid people's eyes-closed imagery was, including images of their own past.⁵

Meaning

LSD makes music feel personally meaningful through 5-HT2A receptors.⁴ The sense that what happened in a session meant something is thought to be part of how it helps.

Safety

Music lowers anxiety,¹⁵ and anxiety predicts harder sessions.¹⁷ Familiar, steady music during the onset can make the difference between bracing and letting go.

Awe

Wonder and transcendence in response to music increase under LSD.³ Mystical-type experiences are among the best predictors of lasting benefit in psilocybin trials.²²

Part 6The colour of sound

Two instruments can play exactly the same note and still sound nothing alike. The difference is timbre, the tone colour of a sound, carried mostly in its overtones and texture. It's how you tell a voice you love from a stranger's in one syllable.

When Imperial researchers scanned people listening to music on LSD, the biggest changes in music-evoked brain activity tracked timbral complexity, more than other acoustic features they measured. People's sense of wonder rose with the brain's response to timbre in the precuneus and the right inferior frontal gyrus. This comes from a preprint of 16 people, so treat it as a lead rather than a fact.²³

If this holds up, it has a practical consequence. We could choose or even compose session music deliberately for its texture, instead of assembling playlists on taste the week before a trial starts. If you want to explore the emotional side of harmony, Why does that chord feel sad? lets you morph a major chord into a minor one and hear the difference.

Part 7The hidden therapist at work

The phrase comes from a 2018 study of 19 people with treatment-resistant depression who received psilocybin with music. Afterwards they were interviewed about the music and rated it. How they experienced the music predicted how much their depression improved a week later, and it predicted mystical-type experience and insight. How intense the drug felt did not predict improvement.¹

Other findings point the same way. In a study of LSD in healthy volunteers, the drug increased the "entropy," or unpredictability, of brain activity, and that increase predicted a rise in the personality trait openness two weeks later. The link was strongest during the scans in which people were listening to music.²⁴ And after psilocybin therapy for depression, people's brains responded more strongly to music than they had before treatment.²⁵

Does the kind of music matter? One small randomized study at Johns Hopkins gave 10 people psilocybin for smoking cessation, alternating a Western classical playlist with an "overtone-based" one built on bowls, gongs and drones. Mystical experience trended a little higher with the overtone music, and challenging experiences were similar. It was too small to settle anything.²⁶ The honest answer is that we don't know yet.

Part 8Shaping a session

Music has been matched to the arc of a session since the LSD therapy programs of the 1960s and 70s. Helen Bonny, a music therapist at the Maryland Psychiatric Research Center, wrote one of the first descriptions of how to do it.²⁷ Most protocols today still follow the same broad shape.

If you run sessions, HELM on this site is a session console and music library built around this arc: you hold the wheel, the music is the wind in the sails.

Part 9How sure are we?

The neuroscience is fascinating and the clinical intuition is strong, but much of the evidence comes from small studies by a handful of groups. A 2026 scoping review found 19 papers from just 10 studies, 330 participants in total. Few compared music with no music, none involved MDMA, and the authors concluded that there is little rigorous data to guide how music should be delivered.⁶ A 2025 review counted only three empirical studies that directly tested music selection.⁷ Here is how I'd grade the main claims in this post.

Part 10Why music is so old

Forty thousand years ago, give or take, someone in what is now southern Germany carved finger holes into a vulture's wing bone. The flutes found in the caves of the Swabian Jura are among the oldest known instruments, older than farming, writing or the wheel.²⁸ Music isn't an extra stapled onto the human story. It has been with us from the start.

Songbirds

Canaries learn new songs each season, and studying them is how scientists first found that adult brains keep making new neurons. Neurons born in adulthood are added to the bird's song-control center.²⁹

Early training

Musicians who started training before about age seven show differences in the fibers connecting the two sides of the brain, suggesting a sensitive period for musical learning.³⁰ In Sesotho, Daniel Levitin notes, one verb, ho bina, means both to sing and to dance.³¹

Moving together

In experiments at Stanford, people who sang or moved in time together then cooperated more in an economic game, even when it cost them.³² Synchrony is social glue.

So many of the wounds we treat in psychiatry have to do with belonging: not being wanted, not fitting, losing the thread to other people. Music is the oldest tool we have for weaving that thread back. Put it next to a medicine that reopens the brain's capacity to change, and you have something very old and something very new in the same room, working on the same person.

Where words fail, music speaks.Hans Christian Andersen

Part 11If you're in the room

If you're the participant

• Ask about the music beforehand. What kind is it, how long is it, and can you hear some of it before the day?
• Say what you love and what you can't stand. A song tied to a painful memory can be swapped out. In the hidden-therapist study, people who disliked the music tended to do worse.¹
• You can ask for a change. Quieter, a pause, or silence for a while are all reasonable requests during a session.
• Notice what the music brings up. Images, memories and feelings that come with a piece are worth talking about in integration. Hearing the music again later can bring the state back.

If you're the clinician

• Treat the playlist as part of the protocol. Write it down, version it, and listen to the whole thing before using it.
• Plan for the onset. Steady, familiar, unhurried music while the drug is coming on, when anxiety tends to run highest.
• Have an alternative ready for people who don't connect with the main playlist, and decide in advance when you'd switch.
• Record what you can. Which tracks played when, and how people rated them afterwards. That's how the field will move past taste.

This post is educational and isn't a guide to using any drug. If you're struggling, talk with a clinician you trust. In a crisis, call or text 988.