Fascia guide
Craniosacral therapy is a very gentle hands-on treatment: a practitioner rests their hands lightly on your body, often your feet, head and the base of your spine, while you lie clothed on a couch. Practitioners say they feel a rhythm in the fluid around the brain and spinal cord and ease restrictions in it. In tests, practitioners didn't agree on that rhythm, and the trials disagree: one review found it eased chronic pain, while two larger reviews found no clear benefit for any condition.
Many people find it relaxing, which UK advertising rules accept as a claim for it. The NHS advises against cranial osteopathy, its close relative, for a crying baby.
This guide explains what happens in a session, what's claimed, what studies of the skull and the rhythm found, and what trials found for pain, headache, pregnancy and babies. It covers the advertising rules, the rare serious harms and who should be careful, with a slow breathing routine you can do yourself, shown in 3D.
You lie clothed while the practitioner's hands rest lightly on you, often starting at the feet. Practitioners say they feel a 'craniosacral rhythm' and ease restrictions in the membranes and fluid around the brain and spinal cord. The idea grew out of cranial osteopathy, first proposed in 1939.
When practitioners felt for the rhythm in the same people, they didn't agree on its rate, and the rates felt at the head and at the feet or sacrum didn't match. One review suggests what they feel is more likely slow natural swings in blood pressure.
In rabbits, pulls like those used in the therapy moved neither the joints between the skull bones nor the pressure inside the skull. In 285 adult skulls, the joints interlocked tightly, and about a third of the skulls had some that had closed.
A 2019 review of 10 trials found it eased chronic pain more than fake treatment for up to 6 months. Two 2024 reviews, of 15 and 24 trials, found no clear benefit for any condition.
The NHS advises against cranial osteopathy for colic: there's little evidence it works, and it may hurt your baby. In trials of manual therapies, the drop in crying disappeared when parents didn't know whether their baby had been treated.
UK advertising rules accept claims about relaxation and well-being, but the regulators have seen no evidence that it treats medical conditions. Serious side effects are rare, but adverse reactions have been reported.
You lie clothed on a treatment couch, or on your side or sitting if that's more comfortable. The practitioner places their hands gently on your body, usually starting at the feet, and moves to other areas, often your head and the base of your spine. Sessions usually last 45 minutes to an hour, and perhaps 30 minutes for babies and small children.
Source 1
A UK training school describes gentle touch that encourages 'subtle movement and fluid flow in the body', like 'gently flushing a sticky pipe', relying on the body's own ability to heal. It teaches students to feel 'the craniosacral rhythm', which it calls 'a vital body system, missing from the text books', and says the therapy can benefit 'pretty much anyone', from newborns to older people. The UK advertising rules sum up the idea as relieving restrictions in the flow of the fluid around the brain and spinal cord.
Some training adds SomatoEmotional Release, which the same school says helps clients release 'deep-seated trauma and long-stuck emotional blockages'. In a session, the practitioner may ask gentle questions to help you explore past experiences; the school stresses that you guide how far that goes.
Cranial osteopathy began with William Garner Sutherland, who proposed in 1939 that the skull bones move in a rhythm, pulled by the membranes around the brain, which are pushed in turn by the fluid around the brain and the brain's own movement. John Upledger wrote the textbook CranioSacral Therapy in 1983 and started his institute in 1985. Schools differ: some describe their approach as biomechanical, others as biodynamic.
They share their roots. Cranial osteopathy is offered by osteopaths, who must be registered by law with the General Osteopathic Council. Craniosacral therapists don't have to be registered by law, but craniosacral therapy is on the voluntary register of the Complementary and Natural Healthcare Council (CNHC), which the Professional Standards Authority accredits.
Advertising rules differ too: osteopaths may claim to treat a set list of conditions, mostly of muscles and joints, such as back pain and sciatica, while claims for craniosacral therapy are limited to relaxation and well-being.
Three physical therapists experienced in the method disagreed about the rhythm's rate in 12 people, and the rates weren't related to heart or breathing rates. Two examiners feeling the head and feet at the same time in 28 adults agreed poorly, and the rates at the head and feet differed; the authors said the results didn't support the theory, or the claim that the motion can be felt reliably.
In 40 healthy adults, each of two examiners was consistent with themselves, but they didn't agree with each other. Two registered osteopaths feeling the head and sacrum at once in 11 people agreed poorly or not at all, and their rates differed. Across these studies, agreement between examiners ranged from below 0 to 0.31, on a scale where 0 is chance and 1 is perfect.
A 1999 review found low agreement between examiners, and a 2016 review of 9 reliability studies concluded that the diagnostic procedures of cranial osteopathy 'are unreliable in many ways'.
After consensus training, three experienced osteopaths agreed moderately to almost perfectly on which skull bones felt 'restricted' in 21 adults. Agreeing on a feeling of restriction is a different question from whether the bones move.
Source 23
A 1997 review found the evidence 'scant and inconclusive', with a potential for very small movement. In anaesthetised cats, raising the fluid volume inside the skull pushed two skull bones slightly apart. A chart review of 12 adults reported small changes in x-ray angles after cranial treatment, with no comparison group.
In 13 rabbits, pulls of 5 to 20 grams, like a craniosacral 'frontal lift', changed neither the gap between two skull bones nor the pressure inside the skull; in one rabbit, the gap first opened, by 0.3 mm, at 500 grams. In 285 adult skulls, 91 (32%) had some closed joints on the outer surface, and the open ones interlocked tightly.
A 2024 review of 28 studies, most of them methodologically weak, says Sutherland's theory 'has become dogma, despite scientific progress refuting it'. It found the leading explanation for the rhythm practitioners feel is slow waves in blood pressure, called Traube-Hering-Mayer waves, driven by the nervous system. That's one review's explanation, and we read only its summary.
Source 3
Very small movements of skull bones under large forces are possible, but practitioners can't agree on the rhythm they say they feel, and the light touch used hasn't been shown to move bones or fluid. A 2006 commentary concluded cranial osteopathy lacks a plausible mechanism. Whatever the therapy does, the evidence doesn't support the mechanism its model describes.
Source 27
For: a 2019 review of 10 trials (681 people with neck and back pain, migraine, headache, fibromyalgia, tennis elbow or pelvic girdle pain) found moderate benefits for pain and disability against fake treatment, lasting to 6 months. Its authors also ran the neck pain trial below.
Against: a 2024 review of 15 trials found no statistically significant or clinically relevant change in pain or disability for headache, neck pain, low back pain, pelvic girdle pain or fibromyalgia. A 2024 meta-analysis of 24 trials (1,613 people) found no significant effect on the main outcomes, and said two positive results in its subgroups shouldn't be read as real benefits. Earlier reviews, in 1999 and 2012, found too little good research to judge.
In 54 people with chronic neck pain, 8 weekly sessions eased pain by 21 points out of 100 more than a light-touch sham at 8 weeks, and by about 17 points at 20 weeks, with no serious side effects. Light touch is a fair sham for light touch, but the practitioners knew which treatment they were giving (our note).
Source 29
Pooling 4 trials, it lowered headache pain by 1.1 points out of 10, a change the authors called clinically unimportant, and didn't change disability. The evidence was very low certainty.
Source 30
In 123 pregnant women, adding it to standard care lowered morning pain a little and helped function, but not evening pain or time off sick. The authors called the effects small and 'clinically questionable'.
Source 31
In 142 children aged 5 to 12, six sessions of cranial osteopathy with registered osteopaths made no significant difference to movement, quality of life, pain or sleep at 6 months, though more parents rated their child's general health as improved (38% against 18%).
Source 32
We found only a chart review and survey of 67 people after concussion, an online survey of views on its use in autism, and a pilot of mixed light-touch therapies in 10 soldiers with post-traumatic stress disorder (PTSD) and a head injury, none with a comparison group. The UK advertising regulator, the Advertising Standards Authority (ASA), upheld a complaint about claims to treat long COVID, saying it had seen no evidence that the therapy worked for it.
On colic, the NHS tells parents not to try gentle pressure on a baby's spine or skull (cranial osteopathy), because there's little evidence it works and it may hurt your baby. UK advertising guidance says the evidence for osteopathy in conditions specific to babies, including colic, is limited or negative.
A Cochrane review of 6 trials (325 babies) of chiropractic, osteopathy and cranial osteopathy found that parents' diaries showed about 1.2 hours less crying a day. In the trials where parents didn't know whether their baby had been treated, the difference wasn't significant, and the authors couldn't reach a firm conclusion. A 2018 review rated a similar drop in crying time as moderate-strength evidence, but found no effect on parents' overall rating of improvement.
One craniosacral trial in 58 babies reported 2.5 to 3.3 hours less crying a day than no treatment, with parents aware of the treatment. A 2024 review found craniosacral therapy wasn't effective for colic, and judged two positive trials in children seriously flawed.
Colic usually gets better by 3 to 4 months. The NHS suggests holding or cuddling your baby, keeping them upright to feed, winding them after feeds, gentle rocking, a warm bath and gentle background noise, and talking to your health visitor. A flat patch on a baby's head usually corrects itself; see a GP if you're concerned about its shape or your baby's head is always tilted to one side.
Doses are starting points, not prescriptions.
Slow breathing, about 6 breaths a minute, for at least 5 minutes, done regularly: the move page's dose, which takes it from the NHS breathing exercise for stress. Once or twice a day, plus whenever you feel anxious, is its editorial starting point. The steps are below.
For ordinary neck pain, which usually lasts a few weeks, the NHS suggests keeping your neck moving, neck flexibility exercises, heat or cold packs, and paracetamol or ibuprofen. Our neck pain and tension headache guides cover the exercise with the best evidence.
Source 40
Book it for relaxation if you enjoy it, which is the claim UK advertising rules accept, rather than to treat a condition. A session or two tells you whether you like it (our view).
Source 2
Steps and dose from the move's page: at least 5 minutes, done regularly, from the NHS breathing exercise for stress, with once or twice a day as its editorial starting point. Relaxation is the claim UK advertising rules accept for craniosacral therapy (Committee of Advertising Practice 2022), and slow breathing is a free way to aim for it. If you feel dizzy, light-headed or tingly, make the breaths smaller; with a lung condition, follow your own team's advice.
Loading the 3D figure…
The movement on the figure: Slow breathing (6 breaths a minute)
Built from the sources:
The UK school quoted above recommends that anyone with a medically diagnosed condition uses the therapy 'with the full knowledge and cooperation of their medical specialist'.
Source 1
Each move page gives its own starting dose; pick the one you'll do every day.
If neck pain or headache drew you to craniosacral therapy, these are moves our neck pain guides use; follow each move page's dose where it gives one, or your physiotherapist's plan.
For relaxation, start with the breathing routine; for a sore neck, the strength routine.
A 6-minute breathing routine: slow breathing, crocodile breathing, alternate nostril breathing, box breathing and 4-7-8 breathing.
6 minBeginner5 movesNo equipmentA 12-minute reset for stressful days: 4 minutes of brisk movement, 3 minutes to loosen your neck, shoulders and back, then 5 minutes of slow breathing.
12 minBeginner10 movesA bench or box, A wallA slow, 10-minute floor routine to unwind before sleep: child's pose, twists, happy baby and slow breathing.
10 minBeginner7 movesFirm cushion, A wallDeep neck holds and shoulder-blade strength in under 20 minutes: one of the best-supported kinds of exercise for a stiff, sore neck.
18 minBeginner11 movesA band, An anchor pointOf the 10 chronic pain trials in one review, the 5 that reported safety found no serious side effects, and minor ones were as common with the comparison treatments. The neck pain trial reported no serious side effects.
In one clinic's records of 55 people with traumatic brain injury, 3 had adverse reactions to cranial treatment, and a 1999 review flagged the same report. After any head injury, get medical advice first.
A healthy 3-month-old girl died after forced deep bending of her neck and spine by someone described as a craniosacral therapist; the case report's authors advise against forced manipulation of the spine in newborns and babies. That was forceful, unlike the light touch the schools teach, but it was done under the therapy's name.
A review of harms in children found 15 serious events, including 3 deaths with various practitioners, one of them a craniosacral therapist, and most involved an underlying condition. In manual therapy studies in babies, minor side effects were recorded in about 7 in every 1,000 babies treated, and a Cochrane review found too little data to judge safety.
The NHS says osteopathy may not be safe if you take blood-thinning medicines or have a raised risk of bleeding or stroke, have bones or joints that are very weak or seriously damaged, or have a nerve condition such as peripheral neuropathy.
Source 16
UK advertising rules say claims to treat conditions that need medical supervision could discourage people from getting essential treatment. Don't stop a prescribed medicine without talking to your doctor, and don't put off seeing a GP.
One trial in 123 pregnant women reported small benefits, and its summary doesn't report side effects. Tell the practitioner you're pregnant (our advice).
Source 31
If your baby has a weak or high-pitched cry, or a cry that doesn't sound like their normal cry.
Source 12
Source 41
If you're being sick, feel dizzy, or take medicine that thins your blood.
Source 41
Signs of a stroke: a drooping face, a weak arm or slurred speech, even if they pass. Other signs include weakness or numbness down one side, blurred vision, confusion and a severe headache.
Source 44
Practitioners feeling the same person don't agree on the rhythm, pulls like those used in the therapy moved nothing in rabbits, adult skull joints interlock tightly, and a 2024 review calls the original theory refuted.
In rabbits, therapy-strength pulls didn't change the pressure inside the skull, and one review suggests the felt rhythm fits slow blood-pressure waves. We found no study in people showing the therapy changes that fluid's flow.
That's taught as SomatoEmotional Release, but we found no study showing emotions are stored in tissue, and the claims for PTSD rest on a 10-person pilot with no comparison group. Our psoas and Rolfing guides reach the same conclusion about stored emotions.
The NHS advises against cranial osteopathy for colic, the drop in crying disappears when parents don't know whether their baby was treated, a 2024 review found it wasn't effective, and a baby has died after forced manipulation under the therapy's name.
We found only surveys and studies without comparison groups. The advertising regulators have seen no evidence that it treats medical conditions, and the ASA upheld a complaint about long COVID claims.
Harm is rare, but adverse reactions have been reported after brain injury, a baby died after forceful manipulation under the therapy's name, and the bigger risk is putting off effective care.
A very gentle hands-on therapy: the practitioner rests their hands lightly on your body, often your feet, head and the base of your spine, while you lie clothed on a couch. Practitioners say they feel a rhythm in the fluid and membranes around the brain and spinal cord, and ease restrictions in it. It grew out of cranial osteopathy.
The reviews disagree. One review of 10 trials found it eased chronic pain more than a fake treatment, but a 2024 review of 15 trials and a meta-analysis of 24 found no clear benefit for any condition. Many people find it relaxing, which is a fair reason to have it.
Only by tiny amounts, if at all. In rabbits, pulls like those used in craniosacral therapy didn't move the joints between the skull bones. In adult skulls those joints interlock tightly, and in about a third of skulls some had closed.
Practitioners describe a slow rhythm they feel through the head and body, said to come from the fluid around the brain. Practitioners feeling the same person don't agree on its rate, and one 2024 review suggests what they feel is more likely the body's slow natural swings in blood pressure.
The NHS advises against cranial osteopathy for a crying baby: there's little evidence it works, and it may hurt your baby. Serious harm is rare, but a baby has died after forced neck manipulation by someone described as a craniosacral therapist. Talk to your health visitor or GP about crying, feeding or head shape.
In trials of manual therapies, parents of treated babies reported about an hour less crying a day, but the difference disappeared in trials where parents didn't know whether their baby had been treated. The NHS advises against cranial osteopathy for colic, which usually gets better by 3 to 4 months.
Pooled trials found a drop in headache pain too small to matter, on very low certainty evidence. Our migraine and tension headache guides cover the options with better evidence.
It's part of some craniosacral training, in which the therapist aims to help you release 'deep-seated trauma' said to be held in the body. We found no study showing trauma is stored in tissue. If strong feelings come up during bodywork, a professional trained in trauma can help, and our exercise and PTSD guide covers ways to move.
They share their roots and ideas. Cranial osteopathy is done by osteopaths, who must be registered by law in the UK, while craniosacral therapists don't have to be registered but can join a voluntary register.
We found only surveys and small studies without comparison groups. The UK advertising regulators have seen no evidence that craniosacral therapy can treat medical conditions, and the Advertising Standards Authority upheld a complaint about claims to treat long COVID. After a head injury, get medical advice first.
In trials it caused no serious side effects, and minor ones were as common as with comparison treatments. Rarely, people with brain injuries have had adverse reactions to cranial treatment. The bigger risk is putting off medical care.
Look for an osteopath on the General Osteopathic Council register, or a craniosacral therapist on the Complementary and Natural Healthcare Council register, which the Professional Standards Authority accredits. Keep seeing your GP about any medical condition.
Movement education, not medical advice. Stop for sharp or increasing pain, numbness or tingling, and get individual advice if you are pregnant, after an injury or surgery, or managing a health condition. Health disclaimer.
