A reflexologist presses a point under the ball of your left foot, pauses, and says there is congestion in your stomach. It is a specific statement about an organ nobody can see, derived from a spot on your skin, and it arrives with confidence.
It is also a claim you can check. Most of what gets offered under the heading of traditional practice makes claims too loose to test. Reflexology is different. Its central idea generates a prediction a researcher can set up and run: two trained practitioners examining the same foot should find the same things, and what they find should match what is wrong with the person attached to the foot.
Two research groups ran that test. One used 76 patients and 13 regions of the body and produced a single number that carries most of the answer.
None of that settles whether an hour of careful attention to your feet does you any good. That is a different question with different evidence, and the useful thing a reader can learn here is how to keep the two apart.
The map, as the practice describes it
Reflexology holds that defined areas on the soles of the feet, and in some schools on the hands and ears, correspond to particular organs and regions of the body. Working a zone is said both to reveal and to influence the state of whatever it maps to. That is the framework as reflexology teaches it, and it should be read as the tradition’s own account rather than as established anatomy.
Proposals for how it might work have come from inside the field. A 2014 systematic review by researchers at the University of Ulster’s Institute of Nursing and Health Research sets them out. An older hemodynamic idea holds that pressure raises blood flow to the matching organ. A nerve impulse idea holds that stimulating a reflex point strengthens a nervous connection to the matching body part. A third holds that the whole effect runs through the autonomic nervous system, the branch that manages heart rate, breathing and blood pressure without conscious input. The Ulster reviewers describe the nerve route as the most promising of the three. None has been confirmed.
One finding from that review deserves saying out loud, because it comes from people sympathetic to the practice rather than from critics. Across the 17 trials they examined, the reviewers counted six distinct styles of reflexology, and they recorded that reflexology maps carry subtle differences from one another, which makes pinpointing a given reflex difficult. The three points they name as most often disputed are the heart, the solar plexus and the pituitary gland.
So the map is not one map. That is true inside the tradition before anyone outside it runs a test.
Three reflexologists, 76 patients, one number
In February 1998, a team led by Anders Baerheim at the University of Bergen published the larger of the two direct tests in Norway’s main medical journal. Three reflexotherapists examined 76 patients about whom they knew nothing. They were asked to locate each patient’s problems from the soles of the feet alone, with no other information. Patient and therapist each then graded problems in 13 different parts of the body.
Agreement between the three therapists was measured with weighted kappa, a statistic that scores how much two raters agree beyond what chance alone would produce. Zero means chance. One means perfect. Across the 13 body regions the values ran from 0.04 to 0.22, and the overall figure was 0.11, with a 95 percent confidence interval of 0.08 to 0.14. For six of the 13 regions the agreement was better than chance at conventional significance.
The authors reported two further details that matter more than the headline. A score built specifically from the colon zone showed no significant difference between patients with many distal intestinal complaints and patients with few. And the errors ran in one direction: the therapists scored higher than the patients did, which is to say they found problems that the patients did not report.
The Bergen authors summed it up in one sentence: the agreement “may be better than pure chance, but is too low to be of any clinical significance.”
That phrasing is more accurate than a flat null, and it is the reason this study is the better of the two to lead with. The reflexotherapists were not scoring at random. They were scoring slightly above random and nowhere near usefully. A paper in Norwegian from 1998 is easy to overlook, and its numbers are more specific than anything written about reflexology since.
Two years later a group at the University of Exeter published a blinded version. Three experienced reflexologists took part. Eighteen adults with one or more of six specified conditions were identified from primary care records, and two reflexologists, blinded to those conditions and monitored while they worked, examined each patient’s feet and rated the probability that each condition was present. The ratings did not vary with whether a condition was there. The receiver operating curves, a standard way of grading how well a test separates sick from well, showed the method performing very poorly. Agreement between examiners was again near nothing.
Both studies are small. Eighteen patients and three practitioners is not a large trial, and 76 patients is modest. They are also the studies that exist, they were designed to address the claim directly, and they agree with each other. On the house scale used across this library, the diagnostic claim sits in the last category: tested, and not supported.
What a government evidence review found about symptoms
Here is where readers usually go wrong. A foot map that does not replicate is not evidence that nothing happened on the table. Those are two independent questions, and the second one has its own literature.
The baseline was set in September 2009, when the Medical Journal of Australia published a systematic review by Edzard Ernst, the retired Exeter professor who has written more critical reviews of complementary medicine than anyone alive. He searched six databases with no language restriction and found 18 randomized trials of reflexology delivered by trained reflexologists. They covered a scattered list of conditions: anovulation, asthma, back pain, dementia, diabetes, cancer, foot swelling in pregnancy, headache, irritable bowel syndrome, menopause, multiple sclerosis, the period after surgery, premenstrual syndrome. Only four of those had more than one trial. Five of the 18 trials produced positive results. Methodological quality, scored on the Jadad scale, was often poor, and sample sizes were generally low.
The sentence in that review that does the real work is the one about quality sorting. Most of the higher-quality trials did not generate positive findings. When a literature behaves that way, the usual explanation is that the positive results are coming from the weaker designs.
Because Ernst is a committed critic, his conclusion on its own would not be enough. What makes it load-bearing is that a completely independent exercise, run fifteen years later by different people under a government contract, reached a compatible answer by a different route.
That exercise was commissioned by the Australian Government Department of Health and Aged Care, overseen by the National Health and Medical Research Council, and carried out by Cochrane Australia. Its reflexology evaluation is dated 14 November 2024. It was built to answer a payer’s question, which is whether the practice should be eligible for private health insurance rebates, so it is neither a promotional document nor a debunking exercise.
The scale is worth stating. The reviewers screened 1,239 records, assessed roughly 380 full reports, and included 174 studies. Of those, 123 went into the evidence synthesis and about 100 contributed to at least one meta-analysis. These were randomized trials, not reviews of reviews.
Their ratings use GRADE, a system that grades how much confidence the evidence supports rather than simply whether a result was statistically significant. The best rating reflexology earned was low certainty, for sleep quality, from 12 trials and 782 participants. Low certainty also attached to fatigue in some chronic conditions and to quality of life in some longer-term conditions. Everything else was very low certainty, including pain, which drew on 46 trials and 3,187 participants. For very low certainty results the reviewers decline to interpret the size of the effect at all.
The reason was consistent across outcomes. Every included study was rated either at high risk of bias or as raising some concerns about it. Not one was clean. The reviewers also recorded likely publication bias and inconsistent results between trials.
The conclusion Cochrane Australia printed is that “it is not possible to draw conclusions about the effects of reflexology with confidence for any condition or outcome.” The review’s summary document notes the low certainty signal on sleep in the same breath, and those two statements are consistent rather than contradictory: under GRADE, low certainty is specifically not confidence.
That conclusion is narrower than it first sounds. It is not a finding that reflexology does nothing. It is a finding that the research cannot tell you, which is a different and more annoying situation, and the reason is the quality of the studies rather than the size of the effects they reported.
The comparison that keeps getting set aside
Almost all of that compares reflexology against nothing in particular. Of the 51 studies the Australian reviewers moved to an evidence inventory rather than the main synthesis, 20 were set aside because they used an active comparator. The headline ratings therefore describe reflexology against no treatment or usual care, not against an ordinary foot rub.
So does the zone work add anything to the rubbing? The Ulster review from 2014 is the only source found here that answers directly, and its answer is the single most useful fact in this article.
Those reviewers examined 17 randomized trials with 697 participants between them, covering 34 physiological or biochemical outcomes such as blood pressure, heart rate, cortisol and salivary amylase. Twelve of the 17 studies showed significant changes within the reflexology group. Only three showed a significant difference between reflexology and the control group, in blood pressure, cardiac index and salivary amylase. Counted as outcome measures rather than studies, the same review gives 11 and four, so the unit changes the number. Overall quality was rated very low.
Then the pattern. Eleven of the 17 trials used a control condition that also involved touching the participant’s feet, whether sham reflexology, a simple foot hold, cream application, non-professional massage or Swedish massage. Those trials showed fewer differences between groups. The trials whose control involved no touching at all showed more. The reviewers drew the obvious inference and wrote it down: touch appears to be an integral part of the response.
That is an honest finding from a sympathetic source, and it does not tell a reader that nothing happens. It tells them what the something is most likely to be.
Blinding is the structural reason this field stays murky. In one of the included studies, two thirds of the control group receiving a plain foot hold correctly identified that they were in the control group. The general problem of designing a credible sham for a hands-on therapy is covered in this library’s article on acupuncture, where the evidence is deepest. The reflexology version is simply that a person being touched can usually tell what kind of touching it is.
Two states license reflexologists, and neither tests them
Consumer writing on this subject tends to say reflexology is unregulated. The truth is more interesting and more instructive.
Washington State licenses reflexologists. The Department of Health requires completion of a state-approved reflexology school or apprenticeship program under chapter 18.108 of the Revised Code of Washington, with a minimum of 200 hours of instruction written into rule and broken down by subject: 30 hours on theory, history, zones, reflex points and contraindications, 40 on body systems, 30 on anatomy and physiology, 5 on business practice and law, at least 25 of supervised practicum, and 70 further homework hours. Applicants must be 18. They must pass a written examination, and that examination belongs to the American Reflexology Certification Board, a private organization. They must also pass an online Washington State jurisprudence examination.
New Hampshire licenses reflexologists too, through a board that also covers structural integrators and Asian bodywork therapists, inside the state’s Office of Professional Licensure and Certification. The core requirement, as the state’s own occupational listing describes it, is a live certificate issued by a national certifying body that the state commissioner has approved, and for reflexology the body named is that same private board. There is no state training requirement. There is no state examination; the listing says testing is whatever the national certifier requires. The fees are 110 dollars to apply and 110 dollars to renew every two years. Reciprocity is not permitted. The document carrying these figures is dated 2021, so a prospective applicant should confirm the current amounts.
Put plainly: New Hampshire’s reflexology license is a private certificate plus a fee. Washington writes the hours into law and then borrows the private body’s test. Both are genuine licenses. Neither is a state assessment of whether this practitioner is any good.
That distinction has practical consequences a consumer can feel. A state that writes 200 hours into rule has set a floor, and a complaint about someone practicing below it has somewhere to go. A state that accepts a private certificate has outsourced both the standard and the judgment, so a complaint about the standard itself is a complaint to a membership organization rather than to a regulator. Neither arrangement is scandalous. They are simply different products wearing the same word.
One more boundary marker. The private board that supplies the competence test in both states is not among the programmatic accrediting agencies the US Department of Education recognizes. That register does cover the bodies that accredit chiropractic, naturopathic, osteopathic, acupuncture and massage programs. It carries no reflexology entry at all.
Beyond those two states there is no reliable count. Whether charging money for reflexology requires a massage license is decided state by state, and no federal agency and no legislative research office publishes a national tabulation; advocacy tallies circulate but carry no method or date. The answer for any particular place has to come from that state’s board.
Safety, stated honestly in both directions
The documented harm record for reflexology is thin and mild. The Australian evaluation turned up no serious safety signal. The Ulster review of 17 trials reported no serious adverse effects in any of them. One trial in that set found that people receiving reflexology reported more discomfort than people receiving a plain foot massage, while both groups rated the experience as pleasant.
Two things are true about that record at once, and a careful article says both. A practice that involves pressure on intact skin genuinely has little scope to injure anyone. And a thin harm record is partly an artifact of nobody collecting harm data systematically, which is a weak foundation for a blanket claim that the practice is safe.
The contraindications that circulate in the practice’s own teaching materials, including deep vein thrombosis, foot ulceration, acute infection and unstable pregnancy, could not be traced here to any named professional body or regulator. They are practitioner tradition, and they should be described that way rather than as findings.
The real risk in this picture is not physical. It is informational. A practitioner who tells you that a tender spot on your arch means something is wrong with an internal organ has made a diagnostic statement that two sets of researchers could not reproduce, and acting on it instead of getting the symptom looked at is how a delay happens.
Some situations belong with a clinician rather than on a table. New or persistent abdominal pain, unexplained weight loss, blood in the stool or urine, a change in bowel habit lasting more than a few weeks, numbness or burning in the feet that is spreading, a foot wound that is not healing, or new swelling and tenderness in one calf all need a medical assessment, and none of them is a reflexology finding. Anyone with diabetes should have foot changes assessed by the clinician who manages their diabetes.
What to ask before you book
Four questions, in the order that matters most.
First: will anyone here tell me what is wrong with my organs based on my feet? If the answer is yes, that specific service failed two direct tests, and you should weigh it accordingly.
Second: what am I buying? An hour of structured, attentive touch has measurable effects on how people feel. The evidence that the zone map adds anything on top of that is what shrinks when the control group also gets its feet handled.
Third: what credential is behind the person, and what did it require? In two states the answer involves a license. In most it does not, and the private certificate is whatever that organization decided it should be.
Fourth: who is managing the symptom I came in about? If the honest answer is nobody, that is the thing to fix first.
The useful summary is narrow and firm. Reflexology’s diagnostic claim has been tested and is not supported. Its symptomatic claims rest on a large but uniformly flawed trial literature, where a government review’s best rating was low certainty for sleep, and where the measured differences tend to shrink when the comparison group’s feet are also being touched. Those are different findings about different claims, and anyone selling or buying reflexology is better served by knowing which is which.
Keep exploring: acupressure.
Sources
- Tidsskrift for Den norske legeforening. “Baerheim A, Algroy R, Skogedal KR, Stephansen R, Sandvik H. "Feet, a diagnostic tool?" Tidsskr Nor Laegeforen, 1998 (76 patients; overall weighted kappa 0.11, 95% CI 0.08 to 0.14). PubMed PMID: 9528375.” pubmed.ncbi.nlm.nih.gov
- Complementary Therapies in Medicine. “White AR, Williamson J, Hart A, Ernst E. "A blinded investigation into the accuracy of reflexology charts." Complement Ther Med, 2000. PubMed. DOI: 10.1054/ctim.2000.0380.” pubmed.ncbi.nlm.nih.gov
- Medical Journal of Australia. “Ernst E. "Is reflexology an effective intervention? A systematic review of randomised controlled trials." Med J Aust, 2009 (18 RCTs; 5 positive). PubMed. DOI: 10.5694/j.1326-5377.2009.tb02780.x.” pubmed.ncbi.nlm.nih.gov
- Evidence-Based Complementary and Alternative Medicine. “McCullough JEM, Liddle SD, Sinclair M, Close C, Hughes CM. "The physiological and biochemical outcomes associated with a reflexology treatment: a systematic review." Evid Based Complement Alternat Med, 2014 (17 RCTs, 697 participants). PubMed. DOI: 10.1155/2014/502123.” pubmed.ncbi.nlm.nih.gov
- Australian Government Department of Health and Aged Care / NHMRC / Cochrane Australia. “Natural Therapies Review 2024: Reflexology Evidence Evaluation, dated 14 November 2024.” health.gov.au
- Australian Government Department of Health and Aged Care / NHMRC. “Natural Therapies Review 2024: Report and Recommendations (copyright 2025; NHMRC process report dated 23 January 2025).” health.gov.au
- Washington State Department of Health. “Reflexologist Certification Requirements (RCW 18.108.070; WAC 246-831).” doh.wa.gov
- New Hampshire Employment Security, Economic and Labor Market Information Bureau. “Licensed, Certified, and Registered Occupations in New Hampshire 2021: Reflexologist.” nhes.nh.gov
- US Department of Education. “Programmatic Accrediting Agencies.” ed.gov