Quitting smoking is famously, brutally hard. Nicotine is one of the most tenaciously addictive substances around, the relapse rates for even the best available treatments are sobering, and almost everyone who has tried to quit, or watched someone they love try, knows the grinding cycle of stopping and starting that defines the experience for so many. So when a small pilot study out of Johns Hopkins, led by Matthew Johnson and published in 2014 in the Journal of Psychopharmacology, reported a smoking abstinence rate that looked almost too good to be true, it got attention. The number was startling. The catch, and there is always a catch, is that the very thing that made the result so eye-catching is also exactly what should make a careful reader slow down.
This article reads the Johnson study with both halves of that tension held firmly in hand. The headline figure is genuinely remarkable, and worth taking seriously. It is also exactly the kind of figure that small, open-label pilot studies are notorious for producing, and understanding why is one of the most useful skills a reader of clinical research can have. So this is partly a story about a promising attack on one of medicine's hardest problems, and partly a worked lesson in why a spectacular number from a tiny uncontrolled study should make you curious rather than convinced. Check the specifics against the original paper before publishing, since the figures here are the whole point.
The problem, and why a new angle was worth trying
Smoking remains one of the largest preventable causes of death in the world, and the difficulty of quitting is not a matter of weak will. Nicotine addiction is a deeply entrenched, self-reinforcing pattern, woven into daily routine and neurochemistry alike, and the existing treatments, nicotine replacement, medications like varenicline, behavioral support, help but leave a great many people relapsing again and again. The long-term success rates for quitting, even with the best available tools, are humbling. Most people who try to quit fail, often many times, before they manage it, if they manage it at all.
That stubbornness is precisely what made psilocybin an interesting thing to try, on the same theoretical logic that drives the depression and alcohol research. If addiction is partly a matter of rigid, locked-in patterns of behavior and thought, the kind of groove that ordinary intention cannot easily climb out of, then a substance that temporarily loosens entrenched patterns and opens a window of psychological flexibility might give a determined person a real shot at rewriting the relationship with the cigarette. The drug would not be doing the quitting. It would be creating an opening that a structured quit attempt could exploit. It is an elegant idea, and smoking is in some ways an ideal place to test it, since the outcome, whether or not someone is still smoking, is concrete and in principle verifiable rather than purely a matter of how a person says they feel.
There was also history humming underneath. The notion that a profound psychedelic experience might break an addiction goes back to the mid-century work on LSD for alcoholism, and the intuition that a single transformative session could shift a stuck behavior had been floating around the field for decades. The Johnson study brought that old idea to one of the most common and deadly addictions there is, dressed in modern methodology. Whether the elegant theory would translate into people actually putting down their cigarettes was the question, and the answer, at least on the surface, was a resounding and almost suspicious yes.
It is worth dwelling on why smoking, specifically, is such a revealing target for this idea, because it sharpens both the promise and the test. Nicotine addiction is unusually woven into the texture of a person's day. It is tied to morning coffee, to work breaks, to driving, to stress, to socializing, a hundred small rituals that each become a cue to light up. That density of conditioning is part of what makes quitting so hard, since the urge is not one habit but dozens of them, triggered by scattered moments throughout ordinary life. The psychedelic hypothesis is interesting here precisely because it does not target any single cue. It aims instead at the whole rigid structure at once, the idea being that a sufficiently profound experience might let a person step outside the entire pattern long enough to see it differently and choose to leave it. If that worked for something as densely conditioned as smoking, it would say something striking about the mechanism, which is part of why the result drew such interest beyond the smoking-cessation world.
The flip side is that smoking also makes an unusually honest test, because the outcome is so checkable. Unlike a mood score, which lives entirely inside self-report, whether someone is smoking can be measured in their breath and blood, in the chemical traces that cigarettes leave behind. That means at least one major source of error, people saying what they think the researchers want to hear, can be largely closed off, which Johnson's team took advantage of. So smoking offers a rare combination, a densely entrenched behavior that is also concretely measurable, which makes it both a demanding target and a relatively trustworthy one. The promise and the rigor are bound together in the choice of what to study.
What they actually did
The study was a pilot, and small, and tellingly it was open-label, which is a phrase that should set off a gentle alarm in any careful reader, for reasons we will get to. It enrolled fifteen participants, all of them long-term smokers who had tried to quit before and failed, the kind of entrenched, treatment-resistant smokers for whom a new approach would matter most. These were not casual smokers looking for a nudge. They were people with a serious, established nicotine habit and a track record of failed attempts.
The intervention combined psilocybin with a structured cognitive behavioral therapy program for smoking cessation. This is the part that is easy to skip past and important not to. The psilocybin was not handed out as a standalone magic bullet. It was embedded in a serious, evidence-based quit program, with the dosing sessions timed around a target quit date and surrounded by the preparation, support, and follow-up that a structured cessation effort involves. Participants received psilocybin in supervised sessions, in the now-familiar comfortable setting with trained monitors, as one powerful element within a broader behavioral treatment. So what was being tested was the combination, psilocybin plus CBT, not psilocybin alone.
And here is the design feature that governs everything that follows. The study was open-label, meaning there was no control group and no blinding. Everyone knew they were getting psilocybin. There was no comparison arm receiving a placebo or an alternative treatment, no group against which to measure what the psilocybin specifically added. The outcome tracked was smoking abstinence, verified not only by self-report but by biological measures, breath and blood markers that confirm whether someone has actually been smoking, which is a genuine strength since it makes the core outcome harder to fake. Participants were followed over time to see how many stayed quit. Keep the open-label, uncontrolled structure firmly in mind, because it is the key that unlocks how to read the spectacular number that came out the other end.
What they found
The headline was extraordinary. At the six-month follow-up, a large majority of participants were confirmed abstinent from smoking, with the figure reported around 80 percent biologically confirmed abstinence at six months. To grasp how startling that is, you have to set it against the baseline. The best conventional smoking-cessation treatments typically produce long-term quit rates in the range of perhaps 30 to 35 percent at best, and often considerably lower. A figure around 80 percent is not a modest improvement on that. It is in a different universe, the kind of number that, taken at face value, would represent a staggering leap over anything else in the field.
Later follow-ups reportedly found that a substantial proportion of participants remained abstinent even longer, at twelve months and beyond, which if anything made the result more remarkable. A treatment that helps people quit and stay quit over a year, when the usual story is relapse after relapse, would be a genuine breakthrough against one of the deadliest habits there is. On its face, the Johnson study looked like it had found something close to a miracle for smoking cessation, and the number duly traveled, cited and re-cited as evidence of psilocybin's extraordinary promise.
Pull quote: An eighty percent quit rate is not a better number than the usual thirty. It is a number from a different planet, and that is exactly the reason to hold it at arm's length until a controlled trial says so too.
Now the discipline kicks in, because that face-value reading is precisely the trap.
The result is genuinely exciting and genuinely worth pursuing. It is also exactly the kind of result that the structure of the study should make you treat with real caution, not because anyone did anything wrong, but because small, open-label, uncontrolled pilot studies systematically tend to produce inflated numbers, and a figure this spectacular from a design this limited is a textbook case for skepticism rather than celebration. The rest of this article is mostly about why.
Why a spectacular number from this design demands caution
Here is the heart of the matter, and it is worth going slowly, because it is one of the most useful lessons in all of research literacy. Several features of this study's design conspire to inflate the apparent effect, and they stack on top of one another.
Start with the absence of a control group. Without a comparison arm, there is no way to know what these particular fifteen highly motivated people would have achieved with the structured CBT program alone, or with any other intensive intervention. Some of them might have quit regardless, carried by their own determination and the serious behavioral support, and with no control group you simply cannot separate the contribution of the psilocybin from the contribution of everything else. The 80 percent is the outcome of the whole package delivered to a particular handful of people, and the design offers no way to attribute it to the drug specifically.
Then there is selection, which bites especially hard here. The participants were not a random sample of smokers. They were people motivated enough to enroll in an experimental psychedelic study to quit smoking, which is an extraordinary level of commitment, and motivation is one of the strongest predictors of quitting there is. A group selected for that much determination might post unusually high quit rates with almost any serious program, before the psilocybin is even considered. Add the open-label expectancy effect, the fact that everyone knew they were receiving a dramatic, much-hyped treatment they had sought out and believed in, and you have a powerful recipe for inflated results, since belief in a treatment can itself drive behavior change, especially for something as psychologically mediated as quitting smoking. And finally, the tiny sample magnifies everything, because with only fifteen people, a couple of participants either way swings the percentage wildly, and small samples are statistically prone to throwing off extreme values that shrink toward something more ordinary when the study is scaled up.
None of this means the result is fake or that psilocybin does nothing for smoking. It means the 80 percent figure almost certainly overstates what a proper controlled trial would find, and that the honest way to read it is as a tantalizing signal generated under conditions that reliably exaggerate, not as a measurement of the drug's true effect. The number is a reason to run the bigger trial, not a result you can trust on its own.
It helps to make the logic concrete with a thought experiment. Imagine you took the same fifteen exceptionally motivated smokers and gave them the identical structured CBT program with no psilocybin at all, just an inert capsule they believed might be active. It is entirely plausible that a meaningful fraction of them would quit anyway, carried by their determination, the intensive support, and their belief in the treatment. Now you have no way of knowing, from Johnson's design, whether the real quit rate without the drug would have been 20 percent or 50 percent, which means you have no way of knowing how much of the 80 percent the psilocybin actually added. The missing comparison is not a minor gap. It is the entire question, and an uncontrolled study leaves it completely unanswered. That is why researchers do not treat even a spectacular open-label result as evidence of efficacy, only as evidence that a controlled trial is worth running.
There is a broader pattern here that is worth internalizing, because it recurs throughout medicine. Early, small, uncontrolled studies of a new treatment very often report dramatic results that then shrink, sometimes dramatically, when the treatment is tested rigorously. This is so common it has a name in the literature, the tendency of effect sizes to decline as studies get larger and better controlled. It is not usually a sign of fraud. It is the predictable consequence of removing, one by one, the factors that inflated the early numbers, the selection, the expectancy, the small-sample noise, the absence of a comparator. A reader who knows this pattern greets a spectacular pilot result not with belief and not with cynicism but with a specific prediction, that the real number, whatever it is, is probably lower, and that the controlled trial will tell us by how much.
Where the study holds up
It would be unfair to treat the study as merely a cautionary tale, because within its scope it was thoughtfully done and made real contributions. The use of biological verification of abstinence was a genuine strength, since confirming with breath and blood markers that participants had actually stopped smoking, rather than relying on their word, removes one major source of error that plagues self-reported behavior change. Whatever else is uncertain, the people counted as abstinent really were not smoking at the time of measurement, which is more than many behavioral studies can say.
Embedding the psilocybin in a structured, evidence-based CBT program rather than testing it as a standalone intervention reflected a sound understanding of how these substances seem to work, as an opening that therapy exploits rather than a magic bullet. The longer-term follow-up, tracking abstinence at six and twelve months rather than just immediately after the quit date, was valuable, since durability is exactly what matters for an addiction and the easiest thing for a study to skip. And the authors, to their credit, understood what kind of study they had run and framed it appropriately as a pilot, a feasibility and signal-detection exercise meant to justify a larger controlled trial, rather than overclaiming that they had proven psilocybin cures smoking. The spectacular number was reported honestly as a preliminary, open-label finding, which is exactly the right framing even if the wider conversation did not always preserve it.
Where it strains
The limits are substantial and mostly already on the table, but worth gathering plainly. The lack of a control group is the central one, since without a comparison arm the study cannot tell you what the psilocybin specifically contributed beyond the CBT, the motivation, and the expectancy. The tiny sample of fifteen means the precise quit rate is statistically fragile, swinging dramatically with each individual outcome and almost certainly destined to shrink in a larger study. The open-label design lets expectancy run unchecked, which matters enormously for an outcome as belief-sensitive as quitting smoking, where conviction that the treatment will work can itself help it work.
The selection of unusually motivated participants limits generalization, since the people who enroll in a psychedelic smoking study are not the average smoker and may post high quit rates with any intensive program. And the blinding problem, which the open-label design does not even attempt to address, means there is no separation at all between drug effect and expectation in this study, by construction.
It is worth being honest that this last point makes the smoking pilot weaker, in one specific respect, than some of the other early psychedelic trials. A study like Grob's cancer pilot at least attempted a placebo and a blind, imperfect as they were. Johnson's design did not try, because as an open-label feasibility study it was not meant to. That is a legitimate choice for an early pilot, whose job is to see whether an approach is worth pursuing at all rather than to measure its true effect. But it does mean that, taken in isolation, the smoking study can say even less about causation than a blinded trial of the same size could. The spectacular number and the maximally permissive design are not unrelated, since the looser the controls, the more room there is for inflating factors to push the number up. A reader should feel the pull of the 80 percent and, in the same breath, notice that it came from the design least equipped to defend it.
Put these together and the picture is clear. The 80 percent figure is a product of a design that systematically inflates, applied to a handful of exceptionally committed people, with no control and no blind. That does not make it worthless, but it does mean it cannot bear the weight that its citation history has often placed on it. It is a signal, a bright and interesting one, generated under conditions that demand it be confirmed before it is believed.
What it justifies, and what it does not
A measured reading lands cleanly between excitement and dismissal. What the Johnson 2014 study justifies is real. It justifies treating psilocybin-assisted smoking cessation as a serious and promising candidate worth rigorous investigation, since even a heavily discounted version of that quit rate would be clinically interesting against the dismal baseline of conventional treatment. It justifies, and indeed motivated, the larger controlled trials needed to find out what the true effect is once the inflating factors are stripped away. And it justifies the general idea that the psychedelic approach to addiction, loosening rigid patterns to enable behavior change, is worth pursuing across different addictions, smoking among them.
What it does not justify is taking the 80 percent at face value as psilocybin's real effect on smoking, because the open-label, uncontrolled, tiny design almost certainly inflates it, probably substantially. It does not justify claims that psilocybin has been shown to be dramatically more effective than existing treatments, since the study had no comparison arm to support any such claim. And it does not justify anyone treating it as proof of anything, since a pilot of this kind proves feasibility and generates a signal, nothing more. The honest reading is that the study found a striking signal under conditions that exaggerate, and that the only way to know what is really there is the controlled trial the pilot was designed to justify.
Why this study still matters
The Johnson 2014 study earns its place for two distinct reasons, one about psilocybin and one about how to read science. On the first, it put psilocybin-assisted smoking cessation firmly on the research map, extending the psychedelic-for-addiction idea to one of the most common and lethal habits in the world, and producing a signal striking enough to drive the field to take the question seriously and fund the larger trials that the pilot called for. Whatever the true effect turns out to be, this study is a large part of why anyone is looking, and that is a real contribution to a problem that kills millions.
On the second, it is close to a perfect teaching case for research literacy, which is why it is worth reading carefully even by people with no particular interest in smoking. The gap between the spectacular headline number and the modest design that produced it is exactly the gap that a careful reader learns to notice, the lesson that an eye-popping result from a small, open-label, uncontrolled pilot is a reason to investigate rather than to believe. Internalize that, and you have a tool you can carry to every breathless health headline you will ever encounter.
It is worth saying plainly that learning this lesson is not the same as becoming a cynic. The point is not to dismiss exciting early results, since dismissing everything is just as lazy as believing everything, and both let you skip the actual work of weighing evidence. The point is to calibrate, to feel genuine excitement about a promising signal while holding, at the same time, a clear-eyed sense of how much that signal can and cannot prove. The Johnson study rewards exactly that double posture, and a reader who can hold both feelings at once, the thrill of the number and the discipline about the design, has learned something more valuable than any single fact about psilocybin.
The honest stance on the study is the one that holds both truths at once. The signal is genuinely exciting and worth chasing hard, and the number is almost certainly too good to be true as stated, and only a controlled trial can tell you how much of it survives. That is not a contradiction. It is just what it looks like to read a promising pilot study like a grown-up.