Homeopathy and its “memory of water” premise are a debunked pseudoscience whose remedies, diluted until no molecule of the original substance remains, work no better than placebo

The dilution problem: nothing left to work with

The first and most decisive objection is not clinical but arithmetic. Matter is made of a finite number of molecules. A given amount of a substance contains a specific, countable quantity of them, on the order of Avogadro's number, roughly 6 followed by 23 zeros of molecules per mole. Dilute a solution far enough and you reach a point where, on average, not a single molecule of the original substance remains in the bottle. For the C scale, that threshold sits at about 12C. Anything more dilute, and Hahnemann prized preparations far beyond it, is overwhelmingly likely to be nothing but water, alcohol, or a sugar pill onto which such liquid was dropped.

A 30C remedy is diluted so far past that limit that, to expect even one molecule of the starting substance, you would need a volume of water vastly larger than the Earth, indeed larger than the solar system. This is not a subtle statistical quibble; it is a basic consequence of the atomic theory of matter that was being nailed down in chemistry at the very time homeopathy was spreading. The remedy, in the strict physical sense, is not there.

Past about a 12C dilution, the odds are that not one molecule of the original substance remains. Hahnemann's favored 30C is diluted trillions of times beyond that point.

Homeopathy also inverts the one relationship that holds across all of real pharmacology: dose response. In every genuine drug, effect tracks concentration; take away the active molecules and you take away the effect. Homeopathy asserts the reverse, that removing more makes it stronger, and offers no measured mechanism for it. That is why the European Academies' Science Advisory Council, in its 2017 review, called the claimed mechanisms “unfounded, implausible and contrary to the established evidence base.” The practice does not sit at the frontier of chemistry; it contradicts its foundations.

The memory of water: the Benveniste affair

Defenders have long understood the dilution problem, and their answer is the idea that gives this file its name: even if the substance is gone, the water remembers it. The claim reached the scientific mainstream, briefly, in June 1988, when the journal Nature published a paper by the French immunologist Jacques Benveniste and colleagues reporting that human white blood cells reacted to an antibody solution diluted so far that no antibody molecule should have remained. If true, it implied water could carry a lasting biological imprint. Nature published it only with a rare editorial caveat, noting that the result, if real, would overturn well-established physics and chemistry.

What happened next became a landmark in the policing of scientific claims. Nature dispatched an investigative team to Benveniste's laboratory: the journal's editor John Maddox, the scientific-misconduct investigator Walter Stewart, and the professional magician and skeptic James Randi, brought precisely because he understood how easily observers fool themselves. Under their supervision the experiment was rerun with proper blinding: samples were coded, labels concealed, procedures documented. The reported effect disappeared.

The team's conclusion, published in Nature under the title “High-dilution experiments a delusion,” was that the original results owed to uncontrolled bias rather than any property of water, and that no serious effort had been made to exclude such error. They did not allege deliberate fraud; they found a lab that had convinced itself of an effect that blinding erased. In the decades since, no rigorous, independently replicated experiment has established a “memory of water.” The physics is equally unforgiving: the hydrogen-bond network of liquid water rearranges itself in trillionths of a second, leaving no stable structure to store a message.

Placebo, harm, and a billion-dollar industry

If it does nothing, why is homeopathy a multi-billion-dollar global business, valued in the tens of billions of dollars and still growing, with deep roots in India, France, Germany, and beyond? The answer lies in the very real things that surround the inert pill. A homeopathic consultation is typically long, attentive, and personalized. Most illnesses people bring to it are self-limiting and get better on their own. Symptoms naturally fluctuate, so people tend to reach for a remedy when they feel worst and then improve regardless. Add the genuine, measurable placebo effect that ritual and expectation produce, and you have a reliable engine for satisfied customers, none of it requiring the remedy to do anything at all.

That is also why “it worked for me” cannot settle the question. Personal recovery is exactly what an ineffective treatment would also produce, which is the whole reason medicine relies on controlled, blinded trials rather than testimonials. The plural of anecdote is not data; it is the raw material that placebo effects are made of.

The harm is mostly indirect but not trivial. Patients can delay or abandon effective care for serious conditions in favor of remedies that contain nothing, and the “homeopathic” label has at times been attached to products that were mislabeled, improperly dosed, or contaminated. Regulators have responded to the marketing rather than the metaphysics: the US Federal Trade Commission said in 2016 that it would hold claims for over-the-counter homeopathic drugs to the same evidence standard as any other health claim, and health systems including the NHS in England moved to stop routine funding, with France ending reimbursement. The verdict on the science is settled. Homeopathy is a debunked mechanism dressed in the vocabulary of medicine; what benefit its users feel comes from attention, expectation, and time, not from a memory held in water.