Using methods that don't involve living animals — cell cultures, organoids, computer models, in vitro testing — instead of animal models wherever possible. Replacement is the highest priority in the 3Rs hierarchy. Advances in tissue engineering, organ-on-chip technology, computer simulation, and human cell-based models are expanding replacement possibilities rapidly. The US FDA Modernization Act 2.0 (2022) removed the legal requirement for animal testing for drug approval, explicitly enabling alternative methods.
Using the minimum number of animals needed to achieve scientific objectives. This involves statistical optimization of study design, sharing of animal tissue and data across researchers, and avoiding duplication of studies that have already been conducted. Reduction directly reduces the total number of animals experiencing research-related harm.
Modifying procedures to minimize pain, distress, and suffering — and to improve animal welfare throughout an experiment's duration. This includes better analgesic use, improved housing and husbandry, refined experimental techniques, and humane endpoints (terminating studies when pain reaches a defined threshold rather than at death). Refinement improvements benefit the welfare of animals that must still be used in research even as replacement and reduction targets are pursued.
Mice and rats are used because of their genetic manipulability, short generation time, and physiological similarities to humans for many research questions. Their welfare needs — complex social structure, burrowing and nesting behaviors, exploration — are often poorly met in standard laboratory housing. Research shows that mice in barren cages show more anxiety-like behaviors and higher stress indicators than those with enriched housing. Group housing, nesting material, and complexity improvements are well-evidenced welfare enhancements with minimal cost.
Zebrafish (Danio rerio) are widely used in genetics and developmental biology research due to their transparency and genetic tractability. Fish welfare in research settings has historically received little attention but is gaining more consideration as evidence for fish sentience has strengthened. EU Directive 2010/63 covers fish (including larval forms beyond a developmental stage), requiring welfare considerations that many research facilities are still implementing.
Despite representing a tiny fraction of research animals numerically, primates receive disproportionate attention due to their cognitive complexity and welfare sensitivity. The US NIH ended chimpanzee research in 2015; EU Directive 2010/63 significantly restricts great ape research and requires strong justification for other primate use. Macaques remain widely used in neuroscience and vaccine development research, raising ongoing welfare concerns around social housing, procedure pain, and individual psychological wellbeing.
Dogs (particularly beagles) and cats are used in toxicology testing, surgical training, and some biomedical research. Their use is particularly ethically contentious given their status as companion animals. Many jurisdictions require special justification for dog or cat research use. Pigs are increasingly used as surgical training models and cardiovascular research subjects due to anatomical similarities to humans.
| Region | Key Regulation | Key Features |
|---|---|---|
| European Union | Directive 2010/63/EU | 3Rs legally required; severity classification; reuse restrictions; great ape ban |
| United States | Animal Welfare Act (excludes mice/rats/birds) | Inspections; pain provisions; IACUCs required at institutions |
| United Kingdom | Animals (Scientific Procedures) Act 1986 | Project-level licensing; cost-benefit assessment; severity limits |
| Australia | Australian Code for Care of Animals | Institutional ethics committees; 3Rs mandatory |
| Canada | CCAC Guidelines | Institutional animal care committees; peer assessment |
In the United States, mice, rats, birds, and fish are explicitly excluded from the Animal Welfare Act — meaning the ~80% of research animals that are mice and rats have no federal legal welfare protections in the US. Their welfare depends entirely on institutional policy and voluntary guidance. This is widely recognized as a significant gap in US research animal protection, and advocates have repeatedly (and so far unsuccessfully) pushed for legislative inclusion of these species.
Microfluidic chips seeded with human cells that replicate the physiology of specific organs — liver, lung, kidney, gut, heart — are becoming increasingly sophisticated alternatives to animal models for drug safety and efficacy testing. Companies including Emulate, Tissuse, and others have developed FDA-qualified organ-on-chip models. These technologies can outperform animal models for human-specific pharmacology while eliminating animal welfare concerns entirely.
Mini-organs grown from human stem cells — brain organoids, intestinal organoids, liver organoids — recapitulate human tissue architecture and function in ways that 2D cell cultures cannot. These models are increasingly used for drug screening and disease modeling, with direct human relevance and no animal welfare cost.
Machine learning models trained on existing toxicology datasets can predict compound toxicity without animal testing. QSAR (Quantitative Structure-Activity Relationship) models, AI-based protein structure prediction (AlphaFold), and virtual human simulations are all reducing the need for animal experiments in specific research domains. The FDA's use of computational approaches in regulatory submissions is expanding.
This landmark US legislation removed the statutory requirement that drug candidates must be tested in animals before human clinical trials, giving the FDA discretion to accept alternative methods. While animal testing remains common in drug development, removing the legal mandate opens the door for wider adoption of alternatives. This was a significant legislative shift driven by years of advocacy from both animal welfare and patient safety perspectives.
One of the most impactful refinement interventions is adopting humane endpoints — pre-defined decision points at which an experiment is terminated to prevent further suffering, rather than continuing to a "natural" endpoint (often death). Properly implemented humane endpoints reduce suffering without compromising scientific outcomes and are increasingly standard in well-regulated research. ARRIVE guidelines and similar reporting frameworks promote transparent reporting of humane endpoints in published research.
Welfare science has demonstrated that laboratory animal housing and husbandry significantly affect welfare outcomes and can affect experimental reproducibility:
Cosmetics animal testing bans represent one of the clearest welfare advocacy successes in the research animal space. The EU banned cosmetics animal testing in 2003 (with a final phase-in for complex systemic effects by 2013). Over 40 countries and several major markets have now banned cosmetics animal testing, demonstrating that replacement alternatives are feasible and that consumer pressure can drive regulatory change.
Laboratory animal welfare has improved substantially since the 3Rs framework was proposed in 1959, and alternatives to animal testing are advancing at an accelerating pace. But major challenges remain: regulatory gaps (particularly for mice and rats in the US), implementation gaps between policy and practice, and the enormous scale of animal use globally. The convergence of welfare advocacy, advancing alternatives technology, and growing regulatory acceptance of non-animal methods creates genuine optimism for continued progress — particularly in the context of FDA Modernization Act 2.0 and advancing organ-on-chip and organoid technologies.