People test on animals primarily because living organisms have complex, interconnected biological systems that no single lab technique can fully replicate. When a drug enters your body, it doesn’t just interact with one type of cell. It passes through your digestive system, gets processed by your liver, enters your bloodstream, and potentially affects your heart, brain, kidneys, and immune system simultaneously. For decades, the only way to observe those whole-body interactions before giving a drug to a human was to give it to an animal first.
That reasoning has driven animal testing since at least 1938, when the U.S. Federal Food, Drug, and Cosmetics Act made it mandatory for every new drug development protocol. The law was a direct response to drug disasters that killed people because no safety testing had been done beforehand. For most of the modern pharmaceutical era, animal studies weren’t just common practice. They were legally required.
What Animal Bodies Can Show That Cell Cultures Cannot
The core scientific argument for animal testing comes down to complexity. Humans and other mammals are organisms where organs perform distinct functions in a tightly coordinated network. Hormones circulate between glands and tissues. The gut microbiome influences immune defense and brain function. Infectious agents spread from one organ system to another in patterns that depend on blood flow, immune response, and tissue barriers all working together.
Cell cultures in a lab dish can model the first two or three levels of biological organization: how individual cells behave, how they communicate with nearby cells, and sometimes how a small cluster of tissue responds to a chemical. But they can’t replicate the full cascade of events that happens when a substance moves through an entire body. A drug might seem perfectly safe when tested on liver cells in isolation, then cause dangerous heart rhythm changes once the liver metabolizes it into a different compound that reaches the heart through the bloodstream. That kind of organ-to-organ interaction only shows up in a whole organism.
Mice share approximately 70 percent of the same protein-coding gene sequences with humans. That overlap is far from perfect, but it’s enough that mice, rats, and other mammals develop many of the same diseases we do and respond to many of the same biological mechanisms. This makes them useful, if imperfect, stand-ins for predicting what might happen in a human body.
Which Animals Are Used and How Often
The vast majority of animals in research are mice and fish. Canadian data from 2022, which reflects patterns seen across most Western research systems, shows mice made up 38.1 percent of all animals used in science that year, with fish accounting for 34 percent. Rats represented 3.9 percent. Nonhuman primates, the category that draws the most public concern, accounted for just 0.1 percent of all animals used.
Mice dominate because they’re small, reproduce quickly, and can be bred with specific genetic profiles that make experiments more consistent. Fish are widely used in developmental biology and toxicology because their embryos are transparent, allowing researchers to watch biological processes in real time. Larger animals like dogs or primates are typically reserved for later-stage safety testing where regulators need data from a species whose physiology more closely mirrors human biology.
What Animal Testing Has Made Possible
The list of medical advances developed through animal research covers most of modern medicine. Vaccines for polio, meningitis, tetanus, and hepatitis all came through animal models. So did insulin for diabetes, inhalers for asthma, antibiotics, and the techniques behind organ transplantation. Therapies for high blood pressure, epilepsy, Parkinson’s disease, kidney disease, and multiple forms of cancer were developed and safety-tested in animals before reaching patients.
More recently, animal studies made it possible to develop both the mRNA and viral vector vaccines for COVID-19 at unprecedented speed. Cancer immunotherapies, including immune checkpoint inhibitors that have transformed treatment for melanoma and lung cancer, were designed and refined using animal models. Harvard Medical School cites animal research as foundational to virtually every major category of modern medical treatment.
The Rules Governing How Animals Are Treated
Animal research operates under an ethical framework called the Three Rs, first proposed in 1959 and now embedded in regulations across most countries. The principles are applied in a specific order of priority.
- Replacement comes first: if a non-animal method can answer the scientific question, animals should not be used at all.
- Reduction is next: when animals are necessary, experiments should be designed to use the fewest animals possible while still producing reliable results.
- Refinement applies to any animals that are used: procedures should minimize pain and distress, through appropriate pain relief, better housing conditions, and less invasive study designs.
In practice, refinement has expanded well beyond the research procedure itself. Modern standards address how animals are housed, handled, and bred, recognizing that poor living conditions cause stress that both harms the animal and compromises the science. Institutions using animals in research typically must have their protocols reviewed and approved by ethics committees before any work begins.
Where Animal Models Fall Short
Despite their widespread use, animal tests are far from perfectly predictive. Molecular and physiological differences between species mean animals can miss dangers that later show up in humans. Drug-induced liver injury is a clear example: it accounts for nearly 13 percent of clinical trial failures, meaning drugs that appeared safe in animals went on to harm people.
A study published in Communications Medicine examined a set of drugs that had passed animal safety testing but later caused liver damage in patients, collectively killing more than 200 people. When researchers tested the same compounds using a human liver-on-a-chip, a microengineered device lined with human liver cells, the chip detected the toxic drugs with 87 percent sensitivity. The animal models that originally cleared those drugs had a sensitivity of zero percent for that same set of compounds. They missed every single one.
This gap highlights a fundamental tension: animal testing catches many problems, but the biological differences between species create blind spots that can have fatal consequences.
Alternatives Are Gaining Legal Ground
The legal landscape shifted significantly in late 2022 when the FDA Modernization Act 2.0 became law. The legislation explicitly authorized the use of non-animal alternatives, including cell-based assays, computer models, and organ-on-a-chip devices, to support new drug applications. It removed the blanket requirement that had been in place since 1938 mandating animal studies for every drug in development.
This doesn’t mean animal testing has ended. The FDA still requires traditional animal toxicity studies for many drug types, including repeat-dose safety studies lasting one to six months for antibody-based therapies. But the agency now has the flexibility to accept alternative methods when they provide equivalent or better safety data. If a drug targets a receptor that only exists in humans, for instance, making animal testing scientifically meaningless for that specific question, the FDA can allow sponsors to substitute human cell-based tests or computer modeling instead.
The shift is practical as much as ethical. Technologies like organ-on-a-chip systems are proving they can outperform animal models for specific types of toxicity prediction. As these tools mature and accumulate validation data, the scientific case for defaulting to animal testing weakens for an expanding range of questions. The trajectory is toward using animals only when no adequate alternative exists, a principle that was always embedded in the Three Rs but is now being backed by both technology and law.