Approximately 100-200 million animals are used in research globally each year. The 3Rs framework — Replace, Reduce, Refine — guides welfare progress, and emerging technologies are accelerating replacement.
Animal experimentation remains central to biomedical research, pharmaceutical development, and safety testing — despite growing ethical debate and technological alternatives. Estimates of global research animal use range from 100-200 million animals annually, with mice, rats, fish, birds, and rabbits comprising the vast majority. The welfare implications of this use depend critically on the severity of procedures, housing conditions, and the implementation of welfare refinements.
The 3Rs framework — Replace, Reduce, Refine — was articulated by Russell and Burch in 1959 and remains the foundational welfare framework for laboratory animal science. Its application has evolved substantially:
Replacement refers to avoiding or replacing the use of animals with alternative methods. Absolute replacement uses non-animal methods entirely; relative replacement uses less sentient species. Progress in replacement technology has accelerated dramatically:
Reduction aims to minimize the number of animals used while obtaining sufficient information. Statistical power analysis to right-size experiments, sharing of control groups, and meta-analysis of existing data all contribute to reduction. The EU requirement (under Directive 2010/63/EU) for harm-benefit analysis before approval has created structural pressure for reduction.
Refinement focuses on minimizing suffering and improving wellbeing for animals that must be used. This includes: improved pain recognition and management, better housing and enrichment, humane endpoints that terminate experiments before maximal suffering occurs, and training of personnel in welfare-sensitive procedures.
Accurate pain recognition in laboratory animals is fundamental to welfare refinement. Key advances in 2025 include:
Mouse Grimace Scale (MGS), Rat Grimace Scale (RGS), and equivalent scales for rabbits and ferrets have been validated and are increasingly used in institutional review. Automated facial recognition systems using deep learning can now detect pain grimaces in real-time — enabling continuous welfare monitoring without observer burden.
Despite evidence for pain in laboratory animals following surgical procedures, analgesic underuse remains documented. Studies in 2024 found that approximately 30-40% of surgical procedures in rodents in published research did not report adequate post-operative analgesia. Professional training requirements and institutional review processes are addressing this gap.
Research on animal models of chronic pain — used to study neuropathic and inflammatory pain conditions — raises particular welfare concerns. These models by design create persistent pain states. Welfare refinements include: regular welfare assessment, enhanced housing enrichment, humane endpoint criteria, and development of non-invasive models.
Mice are the most common laboratory animal. Welfare-relevant factors include: enriched housing (nesting material, shelters), group housing for social species (single housing increases stress and confounds results), appropriate lighting cycles, and minimizing handling stress through positive reinforcement training (mice can be trained to walk voluntarily onto handling tools, reducing capture stress).
Non-human primates (NHPs) receive the most regulatory protection. EU Directive 2010/63/EU restricts great ape use and requires special justification for Old World monkeys. NHP housing requires social grouping, complex environmental enrichment, and foraging opportunities. The movement toward phasing out NHP use in regulatory testing is advancing, with the FDA beginning to accept non-NHP alternatives for some categories of drug testing.
Fish are increasingly used in research (zebrafish as a genetic model system; various species in toxicology). Fish welfare science has advanced dramatically — pain research now provides compelling evidence for pain capacity in teleost fish. Regulatory frameworks for fish welfare in research are catching up to this science, with EU and UK frameworks now covering all fish species including larvae.
In the USA, Institutional Animal Care and Use Committees (IACUCs) review research protocols for welfare justification. In the EU, national competent authorities review applications under Directive 2010/63/EU. UK research is regulated under ASPA (Animals (Scientific Procedures) Act 1986). The quality of review varies substantially between institutions, with research suggesting that protocol review is often more procedural than substantive in some contexts.
The 2025 landscape for animal replacement includes:
Laboratory animal welfare science is making genuine progress — in pain recognition, enriched housing, analgesic use, and replacement technology development. The pace of replacement is accelerating, with regulatory frameworks beginning to reflect the availability of alternatives. However, the absolute number of animals used globally has not significantly declined, partly because biomedical research itself is growing. The goal of eliminating research animal use requires both technological investment and cultural change within the research community.