Regulations and laws The regulations that apply to animals in laboratories vary across species. In the U.S., under the
Animal Welfare Act and the
Guide for the Care and Use of Laboratory Animals (the
Guide), published by the National Academy of Sciences, any procedure can be performed on an animal if it can be successfully argued that it is scientifically justified. Researchers are required to consult with the institution's veterinarian and its
Institutional Animal Care and Use Committee (IACUC), which every research facility is obliged to maintain. The IACUC must ensure that alternatives, including non-animal alternatives, have been considered, that the experiments are not unnecessarily duplicative, and
that pain relief is given unless it would interfere with the study. The IACUCs regulate all vertebrates in testing at institutions receiving federal funds in the USA. Although the Animal Welfare Act does not include purpose-bred rodents and birds, these species are equally regulated under Public Health Service policies that govern the IACUCs. The Public Health Service policy oversees the Food and Drug Administration (FDA) and the Centers for Disease Control and Prevention (CDC). The CDC conducts infectious disease research on nonhuman primates (due to end, according to an announcement in November 2025), rabbits, mice, and other animals, while FDA requirements cover use of animals in pharmaceutical research. Animal Welfare Act (AWA) regulations are enforced by the USDA, whereas Public Health Service regulations are enforced by OLAW and in many cases by AAALAC. According to the 2014 U.S. Department of Agriculture Office of the Inspector General (OIG) report—which looked at the oversight of animal use during a three-year period—"some Institutional Animal Care and Use Committees ...did not adequately approve, monitor, or report on experimental procedures on animals". The OIG found that "as a result, animals are not always receiving basic humane care and treatment and, in some cases, pain and distress are not minimized during and after experimental procedures". According to the report, within a three-year period, nearly half of all American laboratories with regulated species were cited for AWA violations relating to improper IACUC oversight. The USDA OIG made similar findings in a 2005 report. With only a broad number of 120 inspectors, the United States Department of Agriculture (USDA) oversees more than 12,000 facilities involved in research, exhibition, breeding, or dealing of animals. Larry Carbone, a laboratory animal veterinarian, writes that, in his experience, IACUCs take their work very seriously regardless of the species involved, though the use of
non-human primates always raises what he calls a "red flag of special concern". A study published in
Science magazine in July 2001 confirmed the low reliability of IACUC reviews of animal experiments. Funded by the National Science Foundation, the three-year study found that animal-use committees that do not know the specifics of the university and personnel do not make the same approval decisions as those made by animal-use committees that do know the university and personnel. Specifically, blinded committees more often ask for more information rather than approving studies. Scientists in India are protesting a recent guideline issued by the
University Grants Commission to ban the use of live animals in universities and laboratories. On April 10, 2025, the FDA announced a Roadmap to Reducing Animal Testing in Preclinical Safety Studies to reduce and phase out animal testing and promote alternatives such as advanced computer simulations and lab grown human "
organoids" and organ-on-a-chip systems. Over 90% of drugs that appear safe and effective in animals do not go on to receive FDA approval in humans predominantly due to safety and/or efficacy issues. On April 29, 2025, the US NIH announced a new initiative to prioritize human-based research technologies and reduce use of animals in NIH-funded research. On July 8, 2025 at a collaborative FDA & NIH Workshop on Reducing Animal Testing, the NIH announced that NIH will no longer seek proposals exclusively for animal models and that all new NIH funding opportunities moving forward should incorporate language on consideration of NAMs.
Numbers Accurate global figures for animal testing are difficult to obtain; it has been estimated that 100 million vertebrates are experimented on around the world every year, 10–11 million of them in the EU. The
Nuffield Council on Bioethics reports that global annual estimates range from 50 to 100 million animals. None of the figures include invertebrates such as shrimp and fruit flies.The USDA/APHIS has published the 2016 animal research statistics. Overall, the number of animals (covered by the Animal Welfare Act) used in research in the US rose 6.9% from 767,622 (2015) to 820,812 (2016). This includes both public and private institutions. By comparing with EU data, where all
vertebrate species are counted,
Speaking of Research estimated that around 12 million vertebrates were used in research in the US in 2016. A 2015 article published in the
Journal of Medical Ethics, argued that the use of animals in the US has dramatically increased in recent years. Researchers found this increase is largely the result of an increased reliance on genetically modified mice in animal studies. In 1995, researchers at Tufts University Center for Animals and Public Policy estimated that 14–21 million animals were used in American laboratories in 1992, a reduction from a high of 50 million used in 1970. In 1986, the U.S. Congress Office of Technology Assessment reported that estimates of the animals used in the U.S. range from 10 million to upwards of 100 million each year, and that their own best estimate was at least 17 million to 22 million. In 2016, the Department of Agriculture listed 60,979 dogs, 18,898 cats, 71,188 non-human primates, 183,237 guinea pigs, 102,633 hamsters, 139,391 rabbits, 83,059 farm animals, and 161,467 other mammals, a total of 820,812, a figure that includes all mammals except purpose-bred mice and rats. The use of dogs and cats in research in the U.S. decreased from 1973 to 2016 from 195,157 to 60,979, and from 66,165 to 18,898, respectively. 2,960 procedures used non-human primates, down over 50% since 1988. A "procedure" refers here to an experiment that might last minutes, several months, or years. Most animals are used in only one procedure: animals are frequently euthanized after the experiment; however death is the endpoint of some procedures. A 'severe' procedure would be, for instance, any test where death is the end-point or fatalities are expected, whereas a 'mild' procedure would be something like a blood test or an MRI scan. The 3Rs state: • Replacement which refers to the preferred use of methods that do not use animals over methods that use animals whenever possible to achieve the same scientific aims. These methods include computer modeling. • Reduction which means using the minimum number of animals possible and to obtain as much information as possible from the same number of animals. • Refinement which refers to methods that alleviate or minimize potential pain, suffering or distress, enhance
animal welfare for the animals used, and use non-invasive techniques. The 3Rs have a broader scope than simply encouraging alternatives to animal testing, but aim to improve animal welfare and scientific quality where the use of animals can not be avoided. These 3Rs are now implemented in many testing establishments worldwide and have been adopted by various pieces of legislation and regulations. Despite the widespread acceptance of the 3Rs, many countries—including Canada, Australia, Israel, South Korea, and Germany—have reported rising experimental use of animals in recent years with increased use of mice and, in some cases, fish while reporting declines in the use of cats, dogs, primates, rabbits, guinea pigs, and hamsters. Along with other countries, China has also escalated its use of
GM animals, resulting in an increase in overall animal use.
Sources Animals used by laboratories are largely supplied by specialist dealers. Sources differ for vertebrate and invertebrate animals. Most laboratories breed and raise flies and worms themselves, using strains and mutants supplied from a few main stock centers. For vertebrates, sources include breeders and dealers including
Fortrea and
Charles River Laboratories, which supply purpose-bred and wild-caught animals; businesses that trade in wild animals such as
Nafovanny; and dealers who supply animals sourced from pounds, auctions, and newspaper ads.
Animal shelters also supply the laboratories directly. Large centers also exist to distribute strains of
genetically modified animals; the
International Knockout Mouse Consortium, for example, aims to provide
knockout mice for every gene in the mouse genome. In the U.S., Class A breeders are licensed by the U.S. Department of Agriculture (USDA) to sell animals for research purposes, while Class B dealers are licensed to buy animals from "random sources" such as auctions, pound seizure, and newspaper ads. Some Class B dealers have been accused of kidnapping pets and illegally trapping strays, a practice known as
bunching. It was in part out of public concern over the sale of pets to research facilities that the 1966 Laboratory Animal Welfare Act was ushered in—the Senate Committee on Commerce reported in 1966 that stolen pets had been retrieved from Veterans Administration facilities, the Mayo Institute, the University of Pennsylvania, Stanford University, and Harvard and Yale Medical Schools. The USDA recovered at least a dozen stolen pets during a raid on a Class B dealer in Arkansas in 2003. Four states in the U.S.—
Minnesota,
Utah,
Oklahoma, and
Iowa—require their shelters to provide animals to research facilities. Fourteen states explicitly prohibit the practice, while the remainder either allow it or have no relevant legislation. In the European Union, animal sources are governed by
Council Directive 86/609/EEC, which requires lab animals to be specially bred, unless the animal has been lawfully imported and is not a wild animal or a stray. The latter requirement may also be exempted by special arrangement. In 2010 the Directive was revised with
EU Directive 2010/63/EU. In the UK, most animals used in experiments are bred for the purpose under the 1988 Animal Protection Act, but wild-caught primates may be used if exceptional and specific justification can be established. The United States also allows the use of wild-caught primates; between 1995 and 1999, 1,580 wild baboons were imported into the U.S. Most of the primates imported are handled by
Charles River Laboratories or by
Fortrea, which are very active in the
international primate trade.
Pain and suffering was administered to this
common sand frog to induce
anesthesia and death. It is generally accepted that animals can feel pain. The extent to which animal testing causes
pain and
suffering, and the capacity of animals to experience and comprehend them, has been debated. According to the USDA, in 2016 501,560 animals (61%) (not including rats, mice, birds, or invertebrates) were used in procedures that did not include more than momentary pain or distress. 247,882 (31%) animals were used in procedures in which pain or distress was relieved by anesthesia, while 71,370 (9%) were used in studies that would cause pain or distress that would not be relieved.
Bernard Rollin of
Colorado State University, the principal author of two U.S. federal laws regulating pain relief for animals, writes that researchers remained unsure into the 1980s as to whether animals experience pain, and that veterinarians trained in the U.S. before 1989 were simply taught to ignore animal pain. In his interactions with scientists and other veterinarians, he was regularly asked to "prove" that animals are conscious, and to provide "scientifically acceptable" grounds for claiming that they feel pain. some critics continue to question how reliably animal mental states can be determined. However, some canine experts are stating that, while intelligence does differ animal to animal, dogs have the intelligence of a two to two-and-a-half-year old. This does support the idea that dogs, at the very least, have some form of consciousness. The ability of invertebrates to experience pain and suffering is less clear, however, legislation in several countries (e.g. U.K.,
New Zealand, Norway) protects some invertebrate species if they are being used in animal testing. In the U.S., the defining text on animal welfare regulation in animal testing is the
Guide for the Care and Use of Laboratory Animals. This defines the parameters that govern animal testing in the U.S. It states "The ability to experience and respond to pain is widespread in the animal kingdom...Pain is a stressor and, if not relieved, can lead to unacceptable levels of stress and distress in animals." The Guide states that the ability to recognize the symptoms of pain in different species is vital in efficiently applying pain relief and that it is essential for the people caring for and using animals to be entirely familiar with these symptoms. On the subject of analgesics used to relieve pain, the Guide states "The selection of the most appropriate analgesic or anesthetic should reflect professional judgment as to which best meets clinical and humane requirements without compromising the scientific aspects of the research protocol". Accordingly, all issues of animal pain and distress, and their potential treatment with analgesia and anesthesia, are required regulatory issues in receiving animal protocol approval. Currently, traumatic methods of marking laboratory animals are being replaced with non-invasive alternatives. In 2019, Katrien Devolder and Matthias Eggel proposed
gene editing research animals to
remove the ability to feel pain. This would be an intermediate step towards eventually stopping all experimentation on animals and adopting
alternatives. Additionally, this would not stop research animals from experiencing psychological harm.
Euthanasia Regulations require that scientists use as few animals as possible, especially for terminal experiments. Regulations focus on whether particular methods cause
pain and
suffering, not whether their death is undesirable in itself. The animals are euthanized at the end of studies for sample collection or
post-mortem examination; during studies if their pain or suffering falls into certain categories regarded as unacceptable, such as depression, infection that is unresponsive to treatment, or the failure of large animals to eat for five days; or when they are unsuitable for breeding or unwanted for some other reason. Methods of euthanizing laboratory animals are chosen to induce rapid unconsciousness and death without pain or distress. The methods that are preferred are those published by councils of veterinarians. The animal can be made to inhale a gas, such as
carbon monoxide and
carbon dioxide, by being placed in a chamber, or by use of a face mask, with or without prior sedation or anesthesia.
Sedatives or
anesthetics such as
barbiturates can be given
intravenously, or inhalant anesthetics may be used. Amphibians and fish may be immersed in water containing an anesthetic such as
tricaine. Physical methods are also used, with or without sedation or anesthesia depending on the method. Recommended methods include
decapitation (beheading) for small rodents or rabbits.
Cervical dislocation (breaking the neck or spine) may be used for birds, mice, rats, and rabbits depending on the size and weight of the animal. High-intensity microwave
irradiation of the brain can preserve brain tissue and induce death in less than 1 second, but this is currently only used on rodents.
Captive bolts may be used, typically on dogs, ruminants, horses, pigs and rabbits. It causes death by a concussion to the brain. Gunshot may be used, but only in cases where a penetrating captive bolt may not be used. Some physical methods are only acceptable after the animal is unconscious.
Electrocution may be used for cattle, sheep, swine, foxes, and mink after the animals are unconscious, often by a prior electrical stun.
Pithing (inserting a tool into the base of the brain) is usable on animals already unconscious. Slow or rapid freezing, or inducing
air embolism are acceptable only with prior anesthesia to induce unconsciousness. ==Research classification==