In the past few years, at more than a dozen research centers across the globe, adults suffering from post-traumatic stress disorder (PTSD) took MDMA (3,4-methylenedioxymethamphetamine)—the psychedelic commonly known as ecstasy. In cozy rooms with soft light and calming music, individual participants worked through trauma with therapists. Half of the participants took MDMA. The others swallowed placebo pills.

After three such sessions, and additional therapy, many of the 100-plus participants improved, according to the report, published last September (1). In the absence of MDMA and through therapy alone, more than 46% of individuals no longer met the criteria for PTSD. For those who took MDMA, it was more than 71%.

Sign up for PNAS alerts.

Get alerts for new articles, or get an alert when an article is cited.

The study is the latest in a series of clinical trials suggesting that MDMA may safely treat PTSD in many patients. In December, the treatment regimen’s developer, MAPS Public Benefit Corporation, became the first company to request Food and Drug Administration approval for a psychedelic drug.

More and more evidence suggests that psychedelics, such as MDMA, LSD (lysergic acid diethylamide), psilocybin (commonly known as magic mushrooms), and DMT (N,N-dimethyltryptamine; the active ingredient in ayahuasca), can be useful tools—in conjunction with psychotherapy—for easing difficult-to-treat mental health conditions. But as these once-villainized drugs come closer to mainstream medicine, a question remains: How do they work? “We need to know so much more,” says neuroscientist Jennifer Mitchell of the University of California, San Francisco, lead author of the MDMA study.

Scientists don’t know how psychedelics improve mental health or why a patient’s environment seems to influence the effect. They also don’t fully understand what triggers the psychedelic experience commonly known as a “trip”—and they’re unsure whether this trip is necessary for overcoming symptoms.

Researchers and clinicians don’t actually need those answers to administer psychedelics safely. But revealing how psychedelics function could help pharmacologists design more effective treatments. “The drugs work super well for some people, but not as well for other people,” Mitchell says. “I’d like to know how to maximize the therapeutic benefit of the drug. And I’d like to know how to reach those people that right now are unreachable.”

Searching for answers, neuroscientists are probing the effects of psychedelics from the molecular level to the entire brain. Ultimately, they’ll need to investigate aspects of drug use that are rarely explored in detail—how individual differences and even the setting interact with these drugs to shape the psychedelic journey and longer-term healing.

Changing Perspectives

Classic psychedelics share a key feature: They have an impact on serotonin receptors in the brain. Serotonin, a chemical messenger that helps neurons communicate, is especially important for regulating mood. Psilocin (a molecule metabolized from psilocybin), DMT, and LSD stimulate neurons by activating many of the same receptors as serotonin. MDMA, a different form of psychedelic, causes a massive release of the brain’s own serotonin, which then activates receptors.

Serotonin does, of course, activate its own receptors. And common antidepressants increase the amount of serotonin available to stimulate receptors. Yet, somehow, psychedelics acting on this same system can cause dramatic effects, such as hallucinations or waves of empathy. And, it seems in some cases, healing, which can persist long after the immediate effects of the drugs have worn off. Ingrained patterns, such as compulsive substance use or constant negative thinking, can be “upended, interrupted, and really blown apart by these psychedelics,” says psychopharmacologist Albert Garcia-Romeu of Johns Hopkins University (JHU) School of Medicine in Baltimore, Maryland.

Researchers recognized this therapeutic potential in the 1950s and ‘60s, when tens of thousands of volunteers participated in psychedelics studies (2). By today’s standards, the research lacked scientific rigor. By the 1970s, governments cracked down on psychedelic drug use due to safety concerns, and the studies came to an end. Research ramped up again in the 1990s, spurred, in part, by new tools for brain imaging (2). Today, such treatments are moving through clinical trials for many mental health conditions, including substance abuse disorders (3), eating disorders (4), and depression (5).

The current treatments aren’t for everyone, cautions Garcia-Romeu. Studies exclude people with conditions that could make a psychedelic experience risky, such as those with a history of psychosis. But from the data available, psychedelics seem to be safe and potentially effective for a number of conditions, he says. And, yet, exactly why these drugs work, often with effects persisting weeks or months after a single dose, remains unknown.

On Target

Many investigating this mystery start from the beginning—the moment that a psychedelic enters the brain and triggers the serotonin receptor known as 5-HT2A. This receptor seems to be necessary for a trip. When scientists block it in mice or humans, the psychoactive drug effects disappear (6). (Researchers, of course, don’t know whether mice trip the way humans do, nor how their environs affect the experience. But mice do a telltale head twitch when under the influence of psychedelics.)

There’s good reason to suspect that this receptor also underlies the therapeutic benefits of psychedelics. Some researchers believe that some psychological conditions occur because environmental or genetic factors cause neurons in the brain’s prefrontal cortex to atrophy, diminishing this executive brain region’s ability to regulate motivation, fear, and reward (7).

In animal models, common antidepressant drugs, if taken continuously, prompt neurons to branch and establish new synapses with other neurons (8). “They just do this very slowly and on a timescale that correlates with their clinical efficacy,” says chemical neuroscientist David Olson, director of the University of California, Davis Institute for Psychedelics and Neurotherapeutics.

In 2018, Olson reported that psychedelics also cause neurons to branch in the prefrontal cortex of animal models, but at a rapid clip (7). “You see those neurons grow back within 24 hours,” he says. “The other thing that is really remarkable is that the effects last for a very long time.” Other drugs that do not act through serotonin 2A receptors, such as cocaine, can also cause neuron branching, but not necessarily as rapidly, robustly, or as targeted to the specific brain regions where this growth could alleviate depression, says Olson. In February 2023, Olson published findings that could help explain why psychedelics cause neurons to quickly branch in the prefrontal cortex (9). In this brain region, a large portion of 5-HT2A receptors are located inside neurons. The purpose of those seemingly sequestered receptors is unknown, as the neuron’s membrane blocks serotonin from reaching them. But psychedelics can pass through the membrane (9).

Working with rat neurons in cell culture plates, Olson’s team found that activating internal receptors with DMT or psilocin triggers neurons to rapidly branch and form new synapses. When researchers helped serotonin reach the inner receptors by using an electrical current to open holes in the cell membrane, they again saw speedy branching. “The location of the 5-HT2A receptor matters,” Olson says.

Relief Without the Trip

Other receptors may matter, too. In addition to activating serotonin 5-HT2A, each psychedelic targets a suite of other receptors (10). LSD, for example, activates additional serotonin receptors, as well as several dopamine receptors (11). And a single receptor could trigger several cellular pathways within the neuron, depending on the compound that activates it. These variations might explain why some compounds that activate the 5-HT2A receptor cause hallucinations, while others do not, says Jason Wallach, a pharmacologist and medicinal chemist at Saint Joseph’s University in Philadelphia, Pennsylvania.

“All of a sudden, there is no big boss man in this brain governing the show.”

—Robin Carhart-Harris

Wallach and his team are incrementally tweaking psychedelic compounds to make them more specific to particular receptors and cellular pathways. In doing so, Wallach hopes to identify more effective drugs, including some that trigger healing without the trip. He and colleagues recently found, for example, that they can control whether a 5-HT2A-triggering molecule causes head twitches in mice by adjusting its ability to activate one cellular pathway over another—potentially opening the door to finding ways to modulate these drugs (12) “There’s a big debate in the field whether or not the psychoactive effects are necessary,” Garcia-Romeu says. “Myself and others feel that those experiences that happen when people are under the influence are actually an important part of the process.”

Researchers often report links between treatment success and the mystical nature of psychedelic experiences. In a 2015 pilot study, Garcia-Romeu and colleagues found that 12 of 15 cigarette smokers who took psilocybin along with therapy to break their addiction were smoke-free 6 months after treatment (13). The more intensely mystical they rated their experience, the larger their decrease in cigarette cravings reported at the 6-month check-in.

But these are only correlations. It’s possible that isolating the individual neural circuits involved could decouple the trip from other effects. By tweaking the chemical structures of different psychedelics, Olson identified several compounds that, at least in mice, cause neurons to branch without triggering those telltale head twitches (14, 15). His biotech startup, the Boston-based Delix Therapeutics, is now conducting a Phase 1 clinical trial of one of these drugs in the Netherlands. The participants are healthy volunteers, but the aim is ultimately to treat major depressive disorder and treatment-resistant depression (16). “A single clinical trial may not give us a definitive answer,” Olson says. “We may not know for many, many years because it is all about what patient population, what disease indication.”

Olson and Wallach both expect that, for some health conditions, a psychedelic experience could help patients. Wallach speculates that when treating depression, for example, a trip might prove useful. “[People] often talk about meaningful transformative personal insights they have had during the psychedelic experience,” Wallach says, and “how that influenced their outlook and perspective in their everyday life.” Such testimonials are prompting him and his team to make fully psychedelic compounds. But for a condition less tied to one’s mental outlook, such as cluster headaches (17), Wallach questions whether the psychedelic experience is critical. “In that case”, he says, “I think it is hard to make an argument that it’s anything more than just this physiological effect.”

One huge benefit of using psychedelics without prompting hallucinations would be scalability. “To administer [psychedelics] safely, you have to give them under the supervision of a healthcare professional,” Olson says. “That dramatically increases the complexity of the treatment, as well as the cost.”

A Whole-Brain Response

Other researchers are zooming out to study how the brain as a whole changes during a psychedelic trip. One leading theory: Psychedelics temporarily topple the brain’s usual hierarchy (18).

This idea rests on a longstanding cognitive science theory, which holds that the brain’s executive control regions, including the prefrontal cortex, create expectations about surroundings and that these expectations dominate our perceptions. Input from other regions, including the visual and auditory cortices, carries less weight in the grand calculation that yields one’s impression of the world.

The psychedelics theory known as the “relaxed beliefs under psychedelics (REBUS) and the anarchic brain model” posits that a brain under the influence of psychedelics gives less weight to executive control, while allowing more input from elsewhere. “All of a sudden, there is no big boss man in this brain governing the show,” says neuroscientist Robin Carhart-Harris of the University of California, San Francisco, a developer of the theory. This change in command might explain the sensory experience of a trip.

In one recent study, altered brain activity did appear to track with that subjective experience. Carhart-Harris and colleagues monitored the brains of 20 healthy adults before, during, and immediately following DMT use (19). While under the influence of the psychedelic, the brain’s waves of electrical activity, captured by electroencephalogram (EEG), were more irregular, indicating a greater complexity of brain activity, which the researchers theorized could stem from information flowing more freely. The level of complexity for each participant also correlated with the individual’s own ratings of the intensity of the psychedelic experience.

Brain imaging data from the same study, captured via functional MRI (fMRI), suggest that a brain under the influence of DMT breaks down typical brain networks and opens up communication between networks. For example, Carhart-Harris’ team found that, on DMT, the default mode network—a brain network known to remain active during self-reflection and daydreaming—communicated less within itself and more with other brain regions, mirroring similar results from his studies of psilocybin (20) and LSD (21). If the temporary reorganization of the brain’s hierarchy allows for greater flexibility in thought patterns that, with therapy, can be shaped into a more positive outlook over time, Carhart-Harris and others theorize that the REBUS model might also explain mental health improvements lasting for weeks or even months after the psychedelic experience (5, 22).

But the model itself is still a matter of debate in the neuroscience community (23). And researchers don’t know yet whether such lasting cognitive changes are related to physiological brain changes during a trip, such as the branching neurons witnessed in animal models.

Enduring Change

Even as researchers continue to explore the neural mechanisms of psychedelic healing, neuroscientists are studying which changes in mental processes underlie reductions in symptoms. JHU researchers, for example, found that patients with major depressive disorder improved their scores in tests of cognitive flexibility—the ability to switch focus between tasks—even 4 weeks after psilocybin treatment (24). But this study found no correlation between improved cognitive flexibility and reduced symptoms.

“You have to tailor these experiences in a way that we are just not accustomed to doing for other sorts of Western medicine. If the environment feels unsafe or threatening, the experience can go very dark very quickly.”

—Jennifer Mitchell

Cognitive neuroscientist Ceyda Sayalı, of JHU’s Center for Psychedelic and Consciousness Research, is also looking at changes in another measure: cognitive effort avoidance. Everyone avoids cognitive effort sometimes—like using a precalculated tip rather than doing the math, Sayalı says. “For people with major depressive disorder, anxiety, or similar mental illnesses, this kind of effort-avoidance behavior can be very pathological.”

In her previous work with healthy participants, cognitive task seekers were better at reducing activity in their default mode networks (dialing down that daydreaming function), while, at the same time, increasing activity in the frontal parietal network, a necessity for task completion (25). Sayalı is now enrolling patients diagnosed with both major depressive disorder and alcohol use disorder in a study that will combine simple number tasks and fMRI imaging to test whether cognitive effort—and the networks that support it—improves a week after psilocybin use.

If these measures do improve, that still would not explain the underlying reason for the persistent change. One challenge in tying the drugs’ effects to their neural substrates is limited funding. The present wave of psychedelic research has been supported in large part by nonprofits and private donors rather than government grants or large pharmaceutical companies (22). Without substantial financial support, clinical trials are often small and lack neural imaging. Complicating matters, even some of the larger, most high-profile studies, including the recent MDMA trials, have been criticized for methodological flaws, such as failing to truly blind the studies (26–29).

A Personalized Approach

Even if psychedelics can be effective, it’s clear that they don’t work for everyone. Researchers are beginning to explore differences among patients that could predict success. “Everyone has a slightly different organization of their brain,” says neuropsychologist Katrin Preller of the Psychiatric University Hospital Zurich in Switzerland. She’s testing how differences in the way brain networks synch up could predict patient outcomes.

The microbiome, too, may impact psychedelic therapy, in part because bacteria in the gut could influence how the body metabolizes these drugs (30). Garcia-Romeu and colleagues are analyzing how gut microbiome composition influences treatment results.

And then there’s genetics. Some people lack a working version of a gene that codes for an enzyme that metabolizes LSD, for example. For these individuals, trips are longer and more intense, according to a 2021 report by researchers at University Hospital Basel in Switzerland (31).

Increasingly, researchers are taking this sort of personalized medicine approach to many drugs—not just the mind-altering sort. But the psychedelics field has long catered to the individual because of two major additional variables: set and setting.

“Set” is the mindset that someone brings to treatment. “Setting” is the environment where the drug is administered, including everything from music to throw pillows. “You have to tailor these experiences in a way that we are just not accustomed to doing for other sorts of Western medicine,” Mitchell says. “If the environment feels unsafe or threatening, the experience can go very dark very quickly.”

This experiential nature of psychedelics complicates clinical trials. It’s hard to truly “blind” studies. Patients typically know whether they’ve taken a psychedelic, even when researchers offer a lower dose or a nonpsychedelic drug as a placebo (32). So expectations of healing could change people’s outlooks and lead researchers to overestimate the drug’s direct effects. The environment where the drug is administered, which includes the therapists being present, could also sway treatment outcomes in ways that differ between clinics, or even between patients in the same clinic.

Aiming to make these factors more transparent, researchers from Imperial College London recently called on over 70 psychedelics researchers, study participants, and therapists to recommend components of setting that should be reported in clinical trials. The consensus, shared in late summer 2023 at a set and setting workshop in the Netherlands, includes a long list of variables, from whether therapists gave participants verbal instructions to whether flowers were in the room.

The importance of the vast majority of these variables, however, has not been tested experimentally. Instead, therapists rely on their own experience and the shared wisdom of their craft. But that wisdom may not always hold up to scientific scrutiny.

Western classical music, for example, is a standard choice for ambience setting in psychedelics treatments. In a pilot study of psilocybin use for smoking cessation, Garcia-Romeu and colleagues swapped the Western classical soundtrack for one that included instruments such as gongs and a didgeridoo in one of a patient’s two sessions. They then let patients choose between the two soundtracks in a third session. Six of 10 patients selected the non-Western music. And there was no statistically significant difference in treatment outcomes (33).

Mitchell worries that results from recent clinical trials will convince people that the drugs are safe anywhere, when, in reality, researchers are still figuring out under what conditions they actually help. All of this work also continues to take place in the shadow of a complicated history. The psychedelics research community is still recovering from when these drugs lost favor in the 1970s, Mitchell says. Credible, safe treatments, she says, will only come through sound experimental approaches that “evaluate exactly when and how and if and why the drugs are effective.”