Homeopathy: does it actually work?
In short: homeopathy treats people with products so diluted that, in most cases, not a single molecule of the starting substance is left. For it to work, water would have to "remember" what it once touched, which contradicts chemistry and physics. The major independent assessments (Belgium 2010, Australia 2015, European science academies 2017, France 2019) all reach the same conclusion: no effect demonstrated beyond placebo. If people feel better, that is real, but it comes from elsewhere: natural recovery, the placebo effect, an attentive consultation. The real danger is not in the pills, but in the effective treatment you don't take.
Updated 11 November 2026.
Millions of Belgians have let little sugar pills melt under their tongue, for a cold, exam nerves or a baby's teething. Many say it works. And yet the scientific community is almost unanimous: homeopathy has no effect of its own. How can we explain the gap? Here is what we know, without mockery or promises.
The verdict at a glance
| The claim | What studies show | Level of evidence |
|---|---|---|
| "The pills treat colds, flu, stress" | No better than placebo in rigorous trials | High (many trials and independent reviews) |
| "Water keeps a memory of the substance" | Never reproduced under blind conditions; contradicted by the physics of water | High, against |
| "The more diluted, the stronger" | No known mechanism; the opposite of all pharmacology | Theoretical |
| "It works on babies and animals" | Parents' and owners' judgement is influenced by their expectations | Moderate |
| "Harmless" | True for the pill itself; false if it replaces a real treatment | Documented cases |
Homeopathy: how is it supposed to work?
Homeopathy was invented at the end of the 18th century by the German physician Samuel Hahnemann. It rests on two principles.
- Like cures like: a substance that causes symptoms in a healthy person would cure the same symptoms in a sick person. Coffee, which keeps you awake, would therefore treat insomnia.
- Dilution and "potentisation": the substance is diluted again and again, with the vial shaken vigorously at each step. The more diluted, the more powerful the remedy is supposed to be.
Hahnemann drew the first principle from an experiment on himself: he took cinchona bark and felt symptoms similar to malaria, which cinchona treats. Nobody has ever reproduced this observation, and "like cures like" has never been shown to be a general rule.
He deserves credit for one thing: in his day, medicine relied on bloodletting, purging and mercury. A remedy that did nothing was often less dangerous than the official treatments.
Dilutions: do the maths
A 1C dilution (one centesimal, written CH in French-speaking countries) divides the concentration by 100. Each additional C divides it by 100 again.
| Dilution | Dilution factor | What is left |
|---|---|---|
| 1C | 1 in 100 | Still plenty of molecules |
| 6C | 1 in 1,000,000,000,000 | A few molecules per pill, at best |
| 12C | 1 in 10²⁴ | Statistically, not a single molecule |
| 30C | 1 in 10⁶⁰ | Nothing, far beyond the threshold |
The threshold comes from Avogadro's number: one mole of a substance (its reference quantity in chemistry) contains about 6 × 10²³ molecules. Beyond 12C, you have diluted more than there were molecules to begin with.
To picture a 30C dilution, which is very common: even starting from a concentrated solution, you would need a volume of water about a trillion times that of the Earth to hope to find a single molecule of the original substance.
Oscillococcinum, sold for flu-like symptoms, is diluted to 200K (another dilution method, called Korsakovian) from an extract of duck liver and heart. In practice, it contains nothing but sugar. For what really protects you against winter viruses, see our article Cold, flu or Covid: what's the difference?.
Why it cannot work
Supporters of homeopathy know that nothing is left in the vial. Their answer is that water keeps an imprint of the substance. That is where the problems start.
Water has no lasting memory. Water molecules are held together by hydrogen bonds that break and re-form within a few picoseconds, a thousandth of a billionth of a second. No structure lasts long enough to travel from the vial to the patient.
The memory would have to be selective. Even the purest water contains traces of minerals, glass and dissolved gases, far more abundant than the diluted substance. It would have to remember only the intended substance and forget everything else, including everything it has touched over millions of years.
The pills are sugar. Sucrose and lactose beads are soaked in the dilution and then dried. The "memory" would also have to pass from water to sugar once the water has evaporated.
The effect should fall with the dose. In pharmacology, the less of a substance you give, the less it acts. Homeopathy claims the opposite, with no mechanism to explain it.
One could object that science does not understand everything. True. But to believe in homeopathy, you would have to overturn whole areas of chemistry and physics that work perfectly well elsewhere. That would require solid clinical evidence. And it does not exist.
What the studies say
Hundreds of clinical trials have been carried out. The pattern repeats itself: the more rigorous the study (large, double-blind, methods set in advance), the closer the effect of homeopathy gets to that of placebo.
The major independent assessments reach the same conclusion:
- 2005, The Lancet: a comparison of 110 homeopathy trials with 110 trials of conventional medicines shows that, once the best-quality trials are selected, the effect of homeopathy is compatible with placebo.
- 2010, Belgium: the Belgian Health Care Knowledge Centre (KCE) finds no evidence of clinical efficacy beyond placebo.
- 2015, Australia: the National Health and Medical Research Council (NHMRC) examines 176 studies and finds no health condition for which homeopathy has been reliably shown to be effective.
- 2017, Europe: the European Academies' Science Advisory Council (EASAC) states that there is no known disease for which homeopathy has a demonstrated effect beyond placebo.
- 2019, France: the Haute Autorité de santé concludes that efficacy is insufficient. Homeopathic remedies have no longer been reimbursed by French social security since 1 January 2021.
There are favourable meta-analyses (studies that pool other studies), published mainly by researchers linked to the homeopathy world. Their own authors often point to small, heterogeneous studies or a risk of bias. In 2025, a team from the RAND Corporation scrutinised 99 recent clinical studies: about two in three controlled trials had a high or unclear risk of bias, and the expert panel convened for the review, which included homeopathy specialists, found this literature at least somewhat deficient on all the types of validity examined.
In this episode of his series "Tronche de Fake" (in French), Thomas Durand, a biologist with a PhD and creator of the popular science channel La Tronche en Biais, reviews these studies and explains how the efficacy of a treatment is assessed.
Why people still feel better
People who say "it works for me" are not lying. They really do feel better. Several well-documented mechanisms explain why.
| Mechanism | What happens |
|---|---|
| Natural recovery | Colds, stomach bugs, sore throats, nerves: most clear up on their own within a few days. The pill taken in the meantime gets the credit. |
| Regression to the mean | We seek treatment when symptoms are at their worst. They are therefore very likely to improve afterwards, whatever we take. |
| Placebo effect | Believing you are being treated genuinely reduces the perception of pain, fatigue and anxiety. |
| The consultation | A homeopathic consultation is often long, with attentive listening. Being heard does you good, and that is a real lesson for conventional medicine. |
| Memory bias | We remember the times it "worked" and forget the others. |
What about babies and animals? It is true that they do not "believe" in the treatment. But it is parents or owners who judge the result, and their view is influenced by what they expect. Besides, little ones' ailments often clear up on their own too.
The real risks
The pill itself is practically harmless: it is sugar. The risk lies elsewhere.
- Delaying real treatment. A bacterial infection, asthma, diabetes or cancer cannot be treated with sugared water. In Italy in 2017, a 7-year-old boy died of an ear infection that turned into encephalitis, treated only with homeopathy during two weeks of fever. His doctor's conviction was made final in 2025.
- Replacing a vaccine. Homeopathic "vaccines" (nosodes) do not protect. Several health authorities have warned against using them instead of vaccines. For the vaccines that matter this winter, see Flu and Covid: time to get vaccinated.
- Mother tinctures and low dilutions, which may contain active substances (see above).
- The cost, modest per box but real over time.
And in Belgium?
Who uses it? According to the KCE, about 6% of Belgians say they use homeopathy.
Who may practise it? Homeopathy is one of the four non-conventional practices covered by the Colla law of 29 April 1999. Since the Royal Decree of 26 March 2014, in force since 22 May 2014, only doctors, dentists and midwives may practise it, after recognised specific training (theory and an internship with a registered homeopath), and only alongside their own profession, within the limits of their competence: a dentist for care of the mouth, a midwife within her own field. This choice was made so that a health professional can make a diagnosis and spot what needs real treatment.
The products. Homeopathic pills are medicines registered with the Federal Agency for Medicines and Health Products (FAMHP). But most are registered through a simplified procedure provided for by EU law: the manufacturer does not have to prove efficacy, and the box may not show any therapeutic indication. That is why you never read "for colds" on a tube of pills.
Reimbursement. Compulsory health insurance (INAMI/RIZIV) does not reimburse homeopathy. Some health insurance funds (mutualités) contribute through their supplementary insurance, depending on the policy. This is a commercial choice by the funds, not recognition of efficacy.
True or false?
"It's natural, so it's risk-free." False. High dilutions contain nothing, but mother tinctures may contain toxic plants. And the main risk is not getting treated.
"It works on babies and animals, so it can't be a placebo." False. Adults assess the result, and their expectations influence it. Little ones' ailments also often pass on their own.
"It's a medicine, since it's sold in pharmacies." Partly true. It is a medicine in the administrative sense, but most are registered without proof of efficacy and without an indication.
"Studies show it works." Some small studies, yes. But when the most rigorous trials are pooled, the effect disappears.
"Trying it can't hurt." For a cold, probably not. For a serious illness or instead of a vaccine, it can.
Key takeaways
Homeopathy is incompatible with what we know of chemistry and physics, and no rigorous study has shown it to do better than a placebo. People who say they feel better are not wrong about what they feel, but about its cause.
If you take it for a minor passing ailment, it does no harm. But never use it instead of an effective treatment, a vaccine or medical advice. And if what you value most is the time and attention your homeopath gives you, tell your doctor: it is a real need that deserves a real answer.
Going further: the theories put forward to defend it
For two centuries, researchers have tried to explain how a remedy without molecules could work. A review published in 2025 by a Swiss and German team close to homeopathy research identified 216 publications, grouped into 14 families of theories that barely overlap. Its authors acknowledge that no theory is generally accepted. A telling detail: they themselves note, with surprise, that placebo did not emerge as an explanation. Here are the best-known theories.
The vital force (Hahnemann). The remedy would act on an immaterial energy that animates the body. This idea predates modern chemistry and cannot be tested, and therefore cannot be refuted.
Water memory (Jacques Benveniste, 1988). The French immunologist publishes an experiment in the journal Nature: antibodies diluted until they disappear supposedly still act on cells. Nature agrees to publish on condition that it can investigate. A team visits the laboratory to check, including the magician James Randi, an expert in sleight of hand. When the tubes are coded and the researchers do not know which ones contain the dilution, the effect disappears. Nobody has ever reproduced it independently.
Digital biology (Benveniste, 1990s). He went further: the "signal" of a molecule could supposedly be recorded and sent by telephone or over the Internet to "inform" water at a distance. Blind tests, including one funded by the US military, showed nothing.
DNA waves (Luc Montagnier, 2009). The Nobel laureate and co-discoverer of HIV claims that highly diluted DNA emits electromagnetic waves capable of recreating DNA in pure water. This work was judged methodologically weak and was never confirmed by other teams.
Quantum theory. The remedy, the patient and the practitioner would be linked by "quantum entanglement". The word comes from physics, but not its meaning: entanglement is a fragile phenomenon observed in isolated particles, not in a human being at 37 °C. Its own supporters admit it cannot be tested.
Nanoparticles (2010). Indian researchers report finding metal nanoparticles in very high dilutions. Even if confirmed, this would ruin the idea of "memory": there would simply be a tiny dose of product, with no demonstrated pharmacological effect at that level.
Water structure and silica from glass. Clusters of water molecules, or silica released from the vial during shaking, would keep the imprint of the substance. These results have not been reproduced, and the objection remains: why the imprint of the substance, and not that of the glass?
Hormesis. Some toxic substances have the opposite effect at very low doses: this is a real phenomenon. But those low doses still contain billions of molecules. It says nothing about a dilution in which none are left.
When a practice needs fourteen explanations that do not overlap, it is usually a sign that there is nothing to explain.
Test yourself
Sources
Anick, D. J., & Ives, J. A. (2007). The silica hypothesis for homeopathy: physical chemistry. Homeopathy, 96(3), 189-195. https://doi.org/10.1016/j.homp.2007.05.003
Belgian Health Care Knowledge Centre (KCE). (2010). Homéopathie : état des lieux en Belgique (KCE Reports 120B). Brussels: KCE.
Chikramane, P. S., Suresh, A. K., Bellare, J. R., & Kane, S. G. (2010). Extreme homeopathic dilutions retain starting materials: a nanoparticulate perspective. Homeopathy, 99(4), 231-242. https://doi.org/10.1016/j.homp.2010.05.006
Cowan, M. L., Bruner, B. D., Huse, N., Dwyer, J. R., Chugh, B., Nibbering, E. T. J., Elsaesser, T., & Miller, R. J. D. (2005). Ultrafast memory loss and energy redistribution in the hydrogen bond network of liquid H2O. Nature, 434, 199-202. https://doi.org/10.1038/nature03383
Davenas, E., Beauvais, F., Amara, J., Oberbaum, M., Robinzon, B., Miadonna, A., Tedeschi, A., Pomeranz, B., Fortner, P., Belon, P., Sainte-Laudy, J., Poitevin, B., & Benveniste, J. (1988). Human basophil degranulation triggered by very dilute antiserum against IgE. Nature, 333, 816-818. https://doi.org/10.1038/333816a0
Dombrowsky, C., Klein, S. D., Würtenberger, S., Baumgartner, S., & Tournier, A. L. (2025). Mapping the theories and models on the mode of action of homeopathy: a scoping review. Journal of Integrative and Complementary Medicine, 31(11), 937-945. https://doi.org/10.1089/jicm.2024.1007
Durand, T. [La Tronche en Biais]. (2019). Homéopathie : les preuves d'efficacité - Tronche de Fake #4.3 [Video]. YouTube. https://www.youtube.com/watch?v=7Cs-AMjyVbM
European Academies' Science Advisory Council (EASAC). (2017). Homeopathic products and practices: assessing the evidence and ensuring consistency in regulating medical claims in the EU. https://www.easac.eu/publications/details/homeopathic-products-and-practices
European Parliament and Council of the European Union. (2001). Directive 2001/83/EC on the Community code relating to medicinal products for human use, Articles 14 to 16. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32001L0083
Haute Autorité de santé (HAS). (2019, 26 June). Évaluation des médicaments homéopathiques soumis à la procédure d'enregistrement prévue à l'article L.5121-13 du CSP (Transparency Committee opinion). https://www.has-sante.fr/upload/docs/application/pdf/2019-06/homeopathie_pic_avis3_cteval415.pdf
Herman, P. M., Crawford, C. C., Maglione, M. A., Newberry, S. J., Amieux, P. S., Blyden-Taylor, K., Khorsan, R., Prenguber, M., Rice, E., Shollar, A., Tyson, T., Vassighi, N., & Coulter, I. D. (2025). The current state of the quality of homeopathic clinical research. Complementary Therapies in Medicine, 88, 103108. https://doi.org/10.1016/j.ctim.2024.103108
Le Quotidien du Médecin. (2017, 12 June). Homéopathie : quand la confiance défie la science. https://www.lequotidiendumedecin.fr/actus-medicales/sante-publique/homeopathie-quand-la-confiance-defie-la-science
Maddox, J., Randi, J., & Stewart, W. W. (1988). "High-dilution" experiments a delusion. Nature, 334, 287-290. https://doi.org/10.1038/334287a0
Montagnier, L., Aïssa, J., Ferris, S., Montagnier, J.-L., & Lavallée, C. (2009). Electromagnetic signals are produced by aqueous nanostructures derived from bacterial DNA sequences. Interdisciplinary Sciences: Computational Life Sciences, 1, 81-90. https://doi.org/10.1007/s12539-009-0036-7
National Health and Medical Research Council (NHMRC). (2015). NHMRC information paper: evidence on the effectiveness of homeopathy for treating health conditions. http://nla.gov.au/nla.obj-788185806
Open. (2025, 10 January). Bimbo di 7 anni morto per un'otite, condannato in Cassazione il medico che lo curò con rimedi omeopatici. https://www.open.online/2025/01/10/condanna-medico-cassazione-omeopatia-bambino-otite/
Royal Decree of 26 March 2014 on the practice of homeopathy. (2014). Belgian Official Gazette (Moniteur belge), 12 May 2014.
RTBF. (2014). N'importe qui ne pourra plus se déclarer homéopathe. https://www.rtbf.be/article/n-importe-qui-ne-pourra-plus-se-declarer-homeopathe-8267604
RTBF. (n.d.). 4 questions pour comprendre les médecines non conventionnelles en Belgique. https://www.rtbf.be/article/4-questions-pour-comprendre-les-medecines-non-conventionnelles-en-belgique-11536591
Shang, A., Huwiler-Müntener, K., Nartey, L., Jüni, P., Dörig, S., Sterne, J. A. C., Pewsner, D., & Egger, M. (2005). Are the clinical effects of homoeopathy placebo effects? Comparative study of placebo-controlled trials of homoeopathy and allopathy. The Lancet, 366(9487), 726-732. https://doi.org/10.1016/S0140-6736(05)67177-2
Walach, H. (2005). Generalized entanglement: a new theoretical model for understanding the effects of complementary and alternative medicine. Journal of Alternative and Complementary Medicine, 11(3), 549-559. https://doi.org/10.1089/acm.2005.11.549