Every year, millions of mice, rats, and other animals are used to test new medicines before those medicines ever reach a person. This idea has been around for over a century. If a drug is safe and works in an animal, it should be safe and work in a person too. That belief has shaped how new treatments are built and approved almost everywhere in the world. But now we know that idea doesn’t work, and animals are no substitute for humans.

Scientists have known about this problem for years and are now speaking up more openly. Most drugs that pass animal testing still fail once they are tried in humans. Different studies report different numbers, but many put the failure rate above 90 percent. That means for roughly every ten drugs that look promising and safe in a mouse or a rat, only about one will end up helping a real patient. Some researchers say that number gets used out of context, since animal testing is meant to catch problems early, not guarantee a drug will succeed. Nearly everyone in the field agrees the gap between animal results and human results is real, and it is large.

This isn’t a small technical detail. It means patients wait longer for treatments. It means research money gets spent chasing drugs that would never work in people. And for conditions like cancer, Alzheimer’s, and other brain diseases, it may be one reason so few genuine breakthroughs have reached the people who need them most.

Why a Mouse Body Isn’t a Human Body

The core issue is that animals and humans, while similar, are not the same. A disease that shows up in a mouse doesn’t always behave the way it does in a person. Sometimes it doesn’t show up in the animal at all.

Matthias Zilbauer, a doctor and researcher at the Cambridge Stem Cell Institute, put it plainly: “A lot of human diseases either do not occur in animals or occur in a different way because they’re not human. We want tests and models that can tell us which treatments work, and in what patients, and a mouse cannot tell us that.”

A recent study from Stanford looked closely at why dementia treatments that work well in mice keep failing in human trials. The researchers found that most mouse studies test a drug early, before any real signs of disease appear. Human trials almost always involve people who are already sick. Giving a drug to a healthy mouse before a disease starts is a very different test than giving that same drug to a person who already has memory loss or brain damage.

The study also found that most mouse research only used male mice, which means treatments that might work differently in women were never properly tested to begin with. In this sex bias, it is something that, in the past, has been endemic in research, with most results reflecting work with males, not females.

None of this means animal research is worthless. It has led to real, life-saving treatments for vaccines, cancer, and countless other conditions. But it also means animal results have limits that haven’t always been discussed honestly, and those limits matter most for diseases we still don’t know how to cure.

New Tools Built From Human Cells

Because of gaps like these, researchers in the UK are now building tools made directly from human tissue instead of relying only on animals.

In Cambridge, scientists are growing tiny clumps of human organs, called organoids, using cells donated by NHS patients. These clumps, some smaller than a millimeter, can mirror how an actual organ behaves, including how it responds to disease and to new drugs. Researchers are starting with organoids for bowel diseases like Crohn’s disease and ulcerative colitis, along with organoids to study tumors and brain conditions.

Other researchers are building “organ-on-a-chip” devices, small devices lined with real human cells that mimic how a drug moves through the body. One company, VivoSphere, is growing heart cells in small gel spheres to catch dangerous drugs before they ever reach an animal or a patient. As the company’s Yuan Tian explained, “If something is going to fail, there’s a lower risk for the animals and also for the patients.”

The UK government has backed this shift with money and deadlines. Under a new national roadmap, regulatory skin and eye irritation testing on animals is set to end by the close of 2026. Botox strength testing on mice is expected to end by 2027, replaced by lab-based methods that don’t require a live animal at all.

By 2030, the government wants to sharply cut testing on dogs and primates that tracks how drugs move through the body. In total, the UK has committed 75 million pounds to build out these new methods and make it easier for researchers to get them approved for use.

The United States is moving in a similar direction. In 2025, the FDA announced it would begin accepting artificial intelligence models, advanced lab-grown tissue tests, and other non-animal methods in place of some animal studies, starting with a pilot program for a class of drugs called monoclonal antibodies.

Why Animals Still Matter, and Why This Isn’t Simple

None of this means animal testing is about to disappear, and it would be unwise to suggest otherwise. Under UK law, animals cannot be used in research if a working non-animal method already exists to answer the same question. But for many questions, that alternative doesn’t exist yet.

Wendy Jarrett, who leads the group Understanding Animal Research, made this point clearly: animals have been fundamental to the discovery of most medicines available today, and to treating disease in people and animals alike. Edith Heard, director of the Francis Crick Institute, said much the same, noting that animal research remains necessary for complex diseases like cancer and Alzheimer’s, even as her institute works to reduce and replace it wherever possible.

This is also, importantly, a question of animal welfare, not just scientific accuracy. Last year in Britain alone, 2.54 million animal testing procedures were carried out, most involving mice, rats, fish, and birds. A small percentage involved dogs, cats, horses, and monkeys. Reducing that number matters on its own terms, separate from whether animal tests are scientifically reliable.

Groups on both sides, from the RSPCA to the pharmaceutical industry, have welcomed the UK’s new roadmap, which suggests this is one of the rare issues where scientists, drug companies, and animal welfare advocates largely agree on the direction, even if they disagree on how fast to move.

What seems to emerge is not a choice between animals and no animals, but a more honest reckoning with what animal testing can and cannot tell us. As Dr. Juliet Dukes of the charity Replacing Animals in Research put it, tools built from real human tissue have the potential to deliver something animal testing never could: medicine built around the actual patient in front of a doctor, not just the mouse in the lab down the hall.