This post is part of Notice & Comment’s symposium on psychedelics and the law. For other posts in the series, click here.

The expansion of psychedelic medicine has brought neuropsychiatry and pharmaceutical law to a structural, methodological, and institutional crossroads. Federal law requires “substantial evidence of efficacy” (FDA Act §505(d)) and regulations (21 CFR 314.126) mandate that drug approvals be grounded in adequate and well-controlled studies designed to minimize experimental bias and ensure reliable conclusions regarding a drug’s true performance. For more than half a century, the primary tool for isolating causal mechanisms, validating efficacy, and eliminating investigator and participant bias in clinical research has been the randomized, double-blind, placebo-controlled trial (RCT). We know that blinding is critical because biases can be as large or larger than genuine treatment effects. Yet when an investigational agent (i.e., a drug that is under study conditions) possesses unmistakable and profoundly transformative psychoactive properties, it challenges this foundational architecture. We review the history of blinding, distinguish psychological and pharmacological paradigms for the action of psychoactive drugs, and explain why the psychological mechanisms may be “unblindable.” While we accept that patient-blinding will fail in this drug class, we suggest that drug approvals may be tenable provided that objective benefits can be proven with robust rater-blinding that is enforced and documented.

A Historical Looking Glass

This methodological dilemma is not entirely unprecedented in the history of science; it echoes the classical epistemological crisis faced by the French Royal Commission on Animal Magnetism in 1784, as recounted by one of us in his book on blinding. Asked by King Louis XVI to evaluate the clinical assertions of Franz Joseph Mesmer, whose practices in Paris were producing dramatic therapeutic breakthroughs and apparent recoveries among patients suffering from complex neurological conditions, a scientific panel led by Benjamin Franklin and Antoine Lavoisier designed some of the earliest blind methodologies in medical history. The commission recognized that the raw observation of clinical improvement was insufficient to prove a novel physical force, and they demanded to know whether the recoveries were driven by a genuine, objective physical fluid or by the psychological weight of expectation and suggestion. By systematically testing blindfolded patients who were deliberately misled regarding the presence or absence of the purported magnetic fluid, the Franklin Commission demonstrated that therapeutic breakthroughs occurred exclusively when participants believed the treatment was occurring.

Consequently, the commission successfully decoupled clinical outcomes from Mesmer’s theoretical mechanism, establishing the baseline expectation of modern scientific medicine: a compound’s objective biological merit must be strictly isolated from the psychological theater and experimental context of its administration. Perplexingly, although Franklin and Lavoisier persuasively demonstrated its value, nearly two centuries would pass before blinding became routine in clinical trials. Even now, blinding is inconsistently applied, poorly measured for success, and rarely reported.

Placebos—inert substances like sugar or fiber—are an important technology for effective blinding. A challenge for blinding is when the investigational agent causes subjective experiences that cannot be disaggregated from the hypothesized treatment mechanism, making it obvious to the patient if she receives a placebo instead. One important innovation is to apply “active placebos.” Rather than an inert substance like a sugar pill, active placebos, such as low-dose stimulants or niacin, provide subjective side-effects like mild flushing, warming sensations, or mild peripheral stimulation. Yet we have no placebo that can replicate the profound, subjective-experience-altering peak psychoactive effects of an active psychedelic substance treatment arm, leaving the trial vulnerable to expectancy bias and the downstream disappointment of the control group.

A modern parallel to Franklin’s test emerged in August 2024 when the Food and Drug Administration (FDA) rejected Lykos Therapeutics’ New Drug Application for midomafetamine-assisted psychotherapy for post-traumatic stress disorder. Despite phase 3 clinical data indicating statistically significant and clinically meaningful reductions in symptom severity on standardized metrics, approximately 90% of participants in the active treatment arm, and an equally problematic proportion of investigators, accurately identified the allocation. The blinding failure renders it nearly impossible for regulators to determine whether the dramatic drop in symptom scores reflects a true molecular intervention or an unprecedented convergence of high participant expectation, social desirability bias, and investigator allegiance. While the Lykos application suffered from a range of other problems it is still instructive here: unless investigators discover a more robust blinding technique, or FDA lowers its standards, it is not clear how psychedelic drugs have a path towards licensure.

The underlying regulatory gridlock stems from an unresolved schism within neuropsychiatry regarding the true therapeutic mechanism of action of psychedelic substances. For simplicity, these approaches can be called the psychological and pharmacological paradigms.

Psychological Paradigm

The dominant psychological paradigm, which has historically informed clinical psychology and early psychedelic advocacy, asserts that the profound, altered state of subjective consciousness, frequently measured and quantified as a mystical-type experience via validated psychometric assessments, functions as the necessary catalyst for long-term psychological breakthrough for treatment refractory psychiatric conditions. Within this framework, the altered subjective experience serves as the core medicine, rendering the molecule inseparable from the psychotherapeutic environment, the clinical protocol, and the surrounding relational context, commonly referred to as the set and setting. Here, the therapeutic outcome is conceptualized as a co-produced phenomenon where the drug opens a temporary psychological window, and manualized talk therapy guides the patient through the restructuring of their narrative identity.

In this psychological paradigm, it is not clear how blinding could even be possible, as it is conceptually incoherent with the hypothesized treatment mechanism. The psychological paradigm suggests that classic psychedelics function primarily as active super-placebos. Under this conceptual framework, the principal pharmacological action of the molecule may not be to directly correct a specific neurochemical deficit, but rather to significantly increase a patient’s suggestibility, downstream neuroplasticity, and vulnerability to environmental context. By dismantling rigid prior cognitive structures and disrupting top-down mental processing, the drug leaves the patient highly susceptible to the expectations of therapists/investigators, the symbolic architecture of the study room, and cultural hype.

The psychological model embraces profound heterogeneity between patients (set and setting) and their paths through treatment, which may also stymie standard trial protocols and blur the lines between the authority of FDA and the states. Under its foundational enabling statutes, FDA is authorized to regulate drugs in interstate commerce; it possesses no statutory mandate to regulate the practice of medicine, a domain historically and legally reserved for state medical boards. Yet, when evaluating a psychedelic-assisted therapy protocol that includes a psychotherapeutic adjuvant, the molecule is structurally dependent upon a manualized, multi-hour psychological intervention that is not within the purview of FDA oversight. A labelled indication—the approved use of the medical product—may require very substantial directions for how the drug will be utilized clinically, though of course physicians may also use them off-label.

Pharmacological Paradigm

In contrast, the pharmacological paradigm models psychoactive substances through a purely biomedical lens, focusing on their properties as rapid acting neuroplastogens. Neuropharmacologists and molecular biologists have argued that classic psychedelics and entactogens trigger structural and functional neural plasticity at the cellular level by stimulating specific receptor pathways. Most notable is the 5-HT2A serotonin receptor, which downregulates the default mode network and upregulates brain-derived neurotrophic factor, aiding in brain cell growth and repair and supporting neuroplasticity, which is the brain’s ability to form new connections. This model posits that the direct biological repair mechanism—the rapid remodeling of dendritic spines (the bump-like structures that stick out from a neuron’s branches), synaptogenesis (brain cells building new links to each other), and the restoration of circuit-level flexibility (helping a brain network get back its capacity to change and adapt, rather than staying stuck in a fixed pattern)—is entirely sufficient for therapeutic recovery. This model relegates the alteration of subjective experience (the trip) as an unnecessary epiphenomenon.

Here, at least in principle, blinding of the subjective experience should be possible, but there remain large methodological questions about how to do so in practice. One idea is to deliver the drug to patients under anesthesia, so they are not consciously aware of the psychoactive effects. This approach was empirically validated in a triple-blind, randomized, placebo-controlled trial where researchers administered anesthetic ketamine to patients suffering from major depressive disorder while they were fully unconscious under general anesthesia for routine, elective surgical procedures. Because the drug was infused exclusively while the patients were anesthetized, the acute psychoactive effects and dissociative sensations were undetectable by the participants. Upon regaining consciousness, participants had no recollection or perceptual cues to indicate their arm assignment, resulting in blinding where guessing patterns matched random chance. With blinding intact, the significant reduction in depression scores did not statistically separate between the ketamine group and the saline placebo group, as both arms experienced an equivalent reduction in symptom severity. While blinding succeeded, the treatment hypothesis failed, producing a null result.

While this particular drug failed, the anesthesia-masked design could provide a path forward for FDA to evaluate other psychedelics. If an investigational drug is administered to an anesthetized patient and still produces substantial, enduring symptom reductions upon regaining consciousness, studies can definitively decouple the biological mechanism of neuroplasticity from the psychological mechanisms of suggestion and super-placebo expectancy.

Alternatively, it may be possible to engineer non-hallucinogenic versions of the psychedelic compounds. If these molecules can have similar treatment effects without the alteration in subjective experience, then blinding becomes unproblematic.

Of course, both approaches (anesthesia administration and non-hallucinogenic versions) are a bet on the pharmacological hypothesis itself. The ketamine case casts doubt, suggesting that we can solve the blinding problem, but at the cost of treatment. This approach is not useful if it turns out that the psychological paradigm is the true conception of how psychedelics work (if they work at all).  

The Hydraulics of Patient Demand

As the formal federal approval pathways face methodological and jurisdictional silos, political and public pressure have begun shifting the landscape toward alternative legal and constitutional avenues. In the aftermath of FDA’s rejection of MDMA, an organized cohort of combat veteran advocacy groups, representing a population suffering from treatment-resistant PTSD, mobilized to invoke “Right to Try” (RTT) laws for legal access to psychedelic compounds. The federal Right to Try Act (RTT) of 2018, along with its various state-level equivalents, was advanced by advocacy groups and patient populations to bypass FDA’s review process (and its Expanded Access pathway). Under RTT frameworks, eligible patients diagnosed with life-threatening diseases or conditions are granted a statutory permission to access experimental, investigational drugs that have successfully undergone Phase 1 trials. RTT does not, however, require that companies provide products, nor does it bypass the Controlled Substances Act.

Given the limits of Federal RTT, jurisdictions like Oregon and Colorado have also utilized ballot initiatives to create parallel, non-medical, state-regulated licensing frameworks for the cultivation, administration, and adult use of natural psychedelics like psilocybin. These state-level regulatory systems operate entirely outside of FDA’s review process and create an unresolved clash with federal drug control enforcement under the Controlled Substances Act, challenging the limits of the Supremacy Clause. And, these pathways have no mechanism for evidence-generation, so we may never learn whether these products are indeed efficacious.

Reconciliation

If the pharmacological model can generate robust treatment effects without the subjective experience that is typical of classical psychedelics, and thereby navigate the standard RCT pathway for approval, we welcome those molecules as additional treatment options. They may well satisfy some of the growing demand for mental healthcare.

But even if they garner some success, we hypothesize that pharmacological approaches may not exhaust the potential treatment benefits of psychedelics. This is to say that there may be additional treatment benefits available only to patients who do experience their treatment as part of a psychological course of care. Thus, the engineering-solutions typified by the pharmacological paradigm will not fully solve the regulatory and epistemic problem.

While blinding has been a critical epistemological and regulatory innovation to avoid false positives, for the approval of drugs whose only effects are subjective, if applied rigidly, blinding may create false negatives, rejecting drugs whose mechanism of action requires a distinct subjective experience. We do not believe there is a fundamental principle that treatments can only be bona fide if the patient is unaware.

While patient knowledge of treatment allocation may be unavoidable, objective outcomes measured by unbiased outcome-raters can provide confidence that the drug is working. Objective outcomes include return-to-work, actigraphy, healthcare utilization, physiological markers, and informant reports. If a psychiatrist can measure a patient’s wellbeing using standardized scales weeks or months after treatment, such data can be compared to both the patient’s own history (in a paired design) as well as to other patients in a control group. Ideally, such outcome raters should have no expectancy about the efficacy of any treatment and thus should be kept independent from the treatment team. Even more so, it is critical that the outcome raters are blinded. This is where FDA should be insistent, preferring objective outcomes with blinding of raters to be implemented, measured, and reported.

Conclusion

Until clinical trials can reliably isolate chemical effects from the distortions of expectation bias, whether through the widespread implementation of anesthesia-masking designs or the synthesis of non-hallucinogenic plastogen analogs, FDA will continue to find itself stuck at the boundaries of regulating a drug and regulating an experience. But as psychedelic therapy continues to expand and garner even greater cultural enthusiasm, the field can no longer avoid this evidentiary reckoning. Psychedelics are not so exceptional as to fall outside standard scientific validation; but psychiatry remains firmly trapped by the blinding issue, wrestling with the unyielding demand of medical science to develop rigorous, experimental architectures that account for exceptional suggestion and measure objective data. While FDA can condition approval on rater-blinding, it cannot dictate the manualized therapy, which is exactly what makes labeling a psychedelic-assisted-therapy product hard.

Julia S. Etkin is a Ph.D. Student in the Department of Psychiatry and Neurosciences at Charité Universitätsmedizin Berlin.

Dr. Christopher T. Robertson is a Professor of Law and N. Neal Pike Scholar in Health & Disability Law at Boston University School of Law. He is also a Professor of Health Law, Policy & Management in the BU School of Public Health.