A comprehensive guide to animal use in scientific research — how many animals are used, what welfare protections exist, what the 3Rs framework means in practice, and what progress is being made toward replacing animal experiments.
Approximately 192 million animals were used in scientific experiments globally in 2015, according to estimates published in PLOS ONE by Taylor et al. (2019). More recent estimates suggest the number may be declining in some countries but is growing globally. The vast majority of animals used are mice and rats (around 80%), but the total also includes fish, birds, guinea pigs, rabbits, dogs, cats, primates, and other species.
In the United States, the Animal Welfare Act covers most research animals but notably excludes mice, rats, and birds bred specifically for research — the three most commonly used species — meaning they lack the same legal protections as other research animals. The EU's Directive 2010/63/EU provides stronger protections, requiring ethical review of all animal procedures and mandatory application of the 3Rs.
Animal experiments vary enormously in their welfare impact. Some experiments involve no more than a brief blood draw; others involve surgical procedures, the induction of disease conditions, or prolonged psychological stress protocols. Regulatory systems classify experiments by severity (mild, moderate, severe, and non-recovery) to allow proportionate oversight and welfare assessment.
The 3Rs framework, developed by Russell and Burch in their 1959 book "The Principles of Humane Experimental Technique," remains the central organizing principle of laboratory animal welfare. Replace means replacing animal experiments with non-animal alternatives wherever possible. Reduce means minimizing the number of animals used while still achieving scientific objectives. Refine means modifying procedures to minimize suffering and improve welfare for animals that must be used.
Replacement has advanced significantly through development of in vitro methods (cell and tissue cultures), computer modeling, organ-on-a-chip technologies, and human volunteer studies. The EU Reference Laboratory for alternatives to animal testing (EURL ECVAM) validates new non-animal testing methods and promotes their adoption in regulatory testing. Several industries — particularly cosmetics (banned in the EU since 2013) — have made substantial progress in replacing animal tests.
Reduction strategies include better experimental design (using statistical power calculations to use the minimum number of animals required), sharing tissue samples between research groups, and using new imaging technologies that allow repeated measurements from the same animal instead of sacrificing cohorts at different time points.
Refinement encompasses pain management, environmental enrichment, better housing, more humane endpoints (euthanizing animals before they reach severe suffering rather than at the end of the experiment), and training staff in animal behavior to recognize and respond to welfare problems earlier.
Laboratory animal welfare is regulated in most high-income countries, but regulations vary significantly in scope and enforcement. The EU Directive 2010/63/EU is considered among the world's most comprehensive frameworks, requiring mandatory ethical review, severity classification, named animal care and welfare officers, and mandatory application of the 3Rs.
The UK Home Office licenses all animal experiments (called "procedures") and requires a project licence for each research program, which includes a harm-benefit analysis. The UK has invested significantly in developing the 3Rs through the NC3Rs (National Centre for the 3Rs), a government-funded body that has funded over 550 research projects on alternatives and refinements since its founding in 2004.
In the US, the Public Health Service Policy and the USDA Animal Welfare Act create a patchwork regulatory system. Institutions receiving federal research funding must have Institutional Animal Care and Use Committees (IACUCs) that review proposed protocols. However, the exclusion of mice, rats, and birds from AWA protections means most animals lack full regulatory oversight.
In many lower-income countries, laboratory animal welfare regulations are weak or inconsistently enforced, creating welfare disparities in global research. As pharmaceutical and contract research industries move to these regions, ensuring minimum welfare standards internationally becomes increasingly important.
Progress toward replacement is accelerating. Organ-on-a-chip technology — miniaturized devices that replicate the physiology of human organs using human cells — is advancing rapidly, with products now commercially available for liver, lung, kidney, gut, and heart models. These systems can outperform animal models for predicting human drug responses because they use actual human cells in a physiologically relevant environment.
Advances in computational modeling and artificial intelligence are enabling in silico (computer-based) prediction of drug toxicity, pharmacokinetics, and efficacy. The FDA Modernization Act 2.0, passed in the US in 2022, removed the requirement that drugs must be tested in animals before human trials — a regulatory change that could significantly accelerate adoption of alternative methods.
The cosmetics industry has largely achieved replacement of animal testing in the EU, demonstrating that replacement is achievable at industrial scale when there is regulatory pressure and industry commitment. The challenge is extending this model to pharmaceutical testing, where safety concerns create higher bars for alternative method validation.
In the near term, the most important welfare gains may come from refinement: better pain management protocols, improved housing and enrichment standards, and more humane endpoints. Research consistently shows that better welfare conditions also improve the reliability of experimental results — stressed animals produce more variable data — making good welfare scientifically as well as ethically advantageous.