Animal testing persists because of a combination of legal mandates, scientific limitations, and the structure of how drugs, chemicals, and consumer products reach the market. No single cause drives it. Instead, several forces reinforce each other: governments require safety data that historically could only come from animals, whole-organism biology can’t yet be fully replicated in a lab dish, and the funding systems that support medical research have long been built around animal models.
Safety Laws Written After a Tragedy
The most direct cause of animal testing is regulation. In 1938, the U.S. passed the Federal Food, Drug, and Cosmetic Act after a drug called Elixir Sulfanilamide killed over 100 people, many of them children. The product had never been tested for safety in any living organism before it went on pharmacy shelves. That law made it illegal to market any new drug without first proving it was safe, and for decades the only accepted way to generate that proof was testing in animals.
From that point forward, the FDA required animal studies before any experimental drug could move into human clinical trials. A company developing a new medication had to show the drug’s effects in at least one, and often two, animal species whose biological responses were considered predictive of what would happen in humans. The studies had to demonstrate how the body processes the drug, what dose would be effective, and what toxic effects appear at higher doses. This wasn’t optional. It was the legal gateway to testing anything in people.
Similar frameworks exist internationally. The Organisation for Economic Co-operation and Development publishes standardized testing guidelines used across dozens of countries. One common protocol, for example, requires feeding a chemical to rodents daily for 28 days to assess what repeated exposure does to organs, blood, and behavior. These guidelines govern not just pharmaceuticals but industrial chemicals, pesticides, and food additives.
Biology Too Complex for a Petri Dish
Even without legal requirements, scientists would still face a fundamental problem: cells in a dish don’t behave like cells inside a body. Researchers can study individual cells and tissues using lab-grown cultures, and those tools are valuable for early-stage work. But the moment a question involves how multiple organ systems interact, how a drug travels through the bloodstream, or how an infection spreads from one tissue to another, isolated cells can’t provide the answer.
Hormonal regulation is a clear example. A hormone released by the gut can influence brain function, immune response, and metabolism simultaneously. No existing lab model captures that full chain of events. The same is true for studying how gut bacteria shape immune defenses, how cancer metastasizes to distant organs, or how a drug that’s harmless to the liver might damage the kidneys. These are whole-body phenomena, and scientists are still far from being able to predict how a complex organism functions by studying its parts separately. Hypotheses generated from cell studies still need to be validated in a living system, or they remain speculative.
The Scale of Current Testing
In fiscal year 2024, USDA data shows roughly 775,000 animals were used in research across the United States. That number covers dogs, cats, rabbits, primates, pigs, hamsters, guinea pigs, birds, sheep, and other species. It does not include mice and rats bred for research, which are excluded from USDA reporting and likely number in the tens of millions.
Of the 775,000 tracked animals, about 407,000 were used in procedures involving no pain. Another 187,000 experienced pain that was minimized with anesthesia or pain relief. Roughly 55,500 were involved in procedures where pain was not minimized, typically because administering pain relief would have interfered with the experiment’s results. Guinea pigs and hamsters accounted for a disproportionate share of this last category, with nearly 19,000 hamsters and 23,000 guinea pigs in that group. Around 60,000 nonhuman primates were used in total across all pain categories.
How Funding Structures Reinforce the Cycle
Government research grants are a major engine behind animal testing. The National Institutes of Health, the largest public funder of biomedical research in the world, has historically expected grant applicants to include animal studies as part of their proposals. When researchers apply for NIH funding, they must justify why animals are necessary, explain that the fewest number possible will be used, and demonstrate that no alternative approach could answer their research question. Peer reviewers then evaluate whether the proposed animal model is the best tool for the job.
This creates a self-reinforcing loop. Researchers train on animal models, publish using animal data, and design future experiments around the same frameworks. Reviewers who evaluate grants are themselves experienced with animal research. The entire ecosystem of academic promotion, journal publication, and funding success has been built around decades of animal-based methodology. Even when alternatives exist, switching requires retraining, revalidation, and a willingness to use methods that grant reviewers may be less familiar with.
Why Alternatives Haven’t Replaced Animals Yet
Newer technologies are closing the gap, but they haven’t eliminated the need for animals. Organ-on-a-chip devices, which use human cells grown on microchips that mimic organ function, can model how a specific tissue responds to a drug. But each chip must be separately validated for each specific use. A liver chip qualified to test one type of drug, say a small molecule painkiller, doesn’t automatically qualify for testing an antibody therapy or an RNA-based treatment. Each new application requires its own round of proof.
There’s also a consistency problem. Because these chips use cells from real human donors rather than standardized cell lines, results can vary depending on the donor, the lab’s selection criteria, and the way cells are handled. Two labs running the same experiment might get different outcomes simply because their donor cells came from different people. Animal models, for all their limitations, offer a level of standardization that newer platforms are still working to match.
Researchers increasingly recognize that animal results often don’t translate to humans anyway. Species differences in how genes, enzymes, and immune systems function mean that a drug working perfectly in mice may fail completely in people. This mismatch likely contributes to the high failure rate of drugs in clinical trials, where roughly 90% of candidates that looked promising in animals never reach the market. The problem is real, but the alternatives aren’t yet robust enough to fill every gap that animal models currently cover.
Recent Legal Shifts
The legal landscape began changing in December 2022, when President Biden signed the FDA Modernization Act 2.0. For the first time since 1938, this law removed the blanket requirement that drugs be tested in animals before entering human trials. Companies can now use cell-based assays, organoids, organ chips, computer modeling, and AI-based methods to generate preclinical safety data and seek FDA approval to begin human studies.
This doesn’t ban animal testing. It simply means animal data is no longer the only path forward. Companies still have to prove their product is safe, and many will continue using animals because the regulatory track record for alternative methods is still being established. The FDA hasn’t yet published detailed guidance on exactly how to qualify these newer methods for every type of drug, so many companies default to the familiar animal-testing route to avoid regulatory delays. The law opened a door, but walking through it requires building an entirely new evidence base that regulators, scientists, and drug companies all trust.