Does Hypnosis Actually Work? What the Research Shows

Hypnosis works for some problems and not for others, and the dividing line is the condition, not the technique. The NCCIH overview of hypnosis lists what it has been studied for: irritable bowel syndrome, anxiety before medical procedures, menopausal symptoms, hot flashes after breast cancer, headaches, PTSD, pain control and quitting smoking. Being studied is not the same as working, and the results across that list are far from even.

Where is the evidence strongest?

Irritable bowel syndrome. On the NCCIH page, this is the clearest case. NCCIH says there is evidence that gut-directed hypnotherapy can relieve IBS symptoms and improve quality of life. In 2021 the American College of Gastroenterology made a conditional recommendation for gut-directed psychotherapies, hypnotherapy among them, and rated the evidence behind it as very low quality. A 2024 review of gut-directed hypnosis for IBS places it, with cognitive behavioral approaches, among the brain-gut therapies with the most trial evidence. The review adds that European and North American gastroenterology guidelines recommend it as a second-line option, and that for mild IBS it can be used as a first choice if the patient prefers it.

Pain. NCCIH describes “a growing body of evidence” that hypnosis may help manage some painful conditions. That wording is cautious on purpose: some conditions, not all, and as part of care rather than in place of it.

Fear and pain in childbirth. A 2024 systematic review of seven randomized trials found that hypnosis during pregnancy reduced fear and pain in labor and improved how women rated the birth.

Phobias. The NHS lists hypnotherapy among the treatments for phobias, alongside cognitive behavioral therapy and medicines. A listing is not a ranking. It tells you hypnotherapy is a recognized option, not that it outperforms the alternatives.

Menopausal hot flashes. Here the evidence has moved. NCCIH cites a 2015 North American Menopause Society position paper that recommended hypnosis for hot flashes while acknowledging that the favorable evidence was limited. The society’s 2023 nonhormone therapy position statement goes further, recommending clinical hypnosis for hot flashes at its highest evidence level.

The wider picture. A 2024 overview of 20 years of meta-analyses by Rosendahl and colleagues gathered 49 meta-analyses covering 261 primary studies. It reports the most robust evidence for hypnosis in people undergoing medical procedures and in people with pain, while noting considerable overlap of studies between reviews and rating only nine of the 49 meta-analyses as high in methodological quality.

Where is it mixed or weak?

Smoking. The trial evidence is unimpressive. A Cochrane review of 14 studies with 1,926 people found no clear evidence that hypnotherapy beats other quit methods, and concluded that any benefit is “small at most.” The certainty of that evidence was rated low to very low.

Anxiety before procedures. NCCIH calls some results promising but says the overall evidence is not conclusive. A 2022 systematic review of hypnosis for dental anxiety found the most evidence for cognitive behavioral therapy, including single-session formats, and contrasting results for hypnosis in its meta-analysis.

Everyday anxiety, sleep, weight and habits other than smoking. The NCCIH hypnosis page does not weigh in on these. When a website tells you hypnosis is proven for them, ask which review it is relying on, then read what that review actually concluded.

Why do two people get different results from the same session?

People differ a great deal in how strongly they respond to hypnosis, a trait researchers call hypnotizability. In a Stanford imaging study, the team screened 545 healthy volunteers to find 36 who scored consistently high and 21 who scored at the extreme low end. The researchers note that only about 10 percent of people are generally classed as highly hypnotizable.

That spread may explain some of the public confusion. One person describes a vivid, absorbing experience; another felt pleasantly relaxed and nothing more. Both can be reporting accurately. They brought different response profiles to the same chair, and a fair trial of hypnosis should account for that before anyone concludes it “works” or “doesn’t.”

Is it just a placebo?

Expectation can shape how people respond to a treatment, and hypnosis is no exception. What the imaging work adds is narrower. In the Stanford study, three distinct changes in brain activity and connectivity appeared only in the highly hypnotizable group, and only while they were hypnotized. That supports the idea that hypnosis has measurable effects in the brain, at least in highly hypnotizable people; a single study of that kind does not establish one brain signature of hypnosis.

That finding does not settle the clinical question. Whether a treatment helps IBS or labor pain more than expectation alone is answered by controlled trials, not by brain scans. The scans show something happens; the trials show whether it helps.

How should you read a bold claim about hypnosis?

Three questions separate a useful finding from marketing:

1. Compared with what? No treatment at all is an easy bar. Another real treatment, such as counseling for smoking, is the bar that matters.
2. Measured how? Pain scores, symptom diaries and confirmed quit rates count for more than “felt better.”
3. For how long? An effect at the end of a session is different from an effect three or six months later.

When a review calls its evidence “low” or “very low” certainty, it is not saying the treatment failed. In the smoking review, Cochrane gives its reasons plainly: there was not enough information, and many of the studies had design problems. A grade like that is a reason to hold the conclusion loosely in both directions.

Is it safe to try?

Generally yes, with clear exceptions. WebMD’s medically reviewed summary says side effects occur in rare cases and can include anxiety, headaches, dizziness or drowsiness. It flags hypnosis as potentially dangerous for people with hallucinations, delusions, drug or alcohol misuse, or schizophrenia, and warns that when it is used to bring back memories by someone who is not a trained professional, it could create fake memories. Pain deserves a medical check before hypnosis, so that a treatable cause is not missed.

Deciding whether it fits your problem

Match the decision to the condition. For IBS, menopausal hot flashes, fear and pain around childbirth, and some kinds of pain, hypnosis has research support as an addition to medical care, and it is reasonable to try it with a clinician who knows the condition. For quitting smoking, treat it as an optional extra beside methods with stronger evidence, not as the plan itself. For anything that touches safety, such as new or changing pain, a serious mental health condition, or a symptom no one has diagnosed, get assessed first.

Then define success before the first session: fewer bad IBS days a month, a calmer response to one specific trigger, a quit date that holds. A concrete target is the only fair way to judge whether the sessions earned their cost.

This article is general information, not medical or psychological advice; talk with a licensed clinician about your own situation.