Mice and rats account for the vast majority of animals used in laboratory research worldwide, far outstripping every other species combined. Their dominance is so pronounced that when researchers say “animal model,” they usually mean a rodent by default. But the full picture includes a surprising range of creatures, from tiny zebrafish and fruit flies to dogs, rabbits, non-human primates, and even chicken embryos, each chosen for specific biological traits that make them suited to particular kinds of questions.
Why Mice and Rats Dominate
Mice and rats have been the go-to species for biomedical research for well over a century. Their popularity comes down to a combination of practical and biological advantages: they are small, relatively inexpensive to house, breed quickly, and have short life cycles that let researchers observe multiple generations in a manageable timeframe. More importantly, they share a striking degree of anatomical, physiological, and genetic similarity with humans, which makes findings in rodents at least plausible starting points for understanding human biology.1PubMed Central. The Mighty Mouse: the impact of rodents on advances in biomedical research
Mice, in particular, benefit from decades of genetic work. Thousands of inbred strains exist, each with a well-characterized genetic background that allows researchers to control for variation in ways that would be impossible with most other species. Knockout mice, in which a specific gene has been deliberately disabled, have become a standard tool for studying gene function and modeling human diseases like cancer, diabetes, and Alzheimer’s. Rats fill a complementary niche. They are common models in toxicology, psychology, and behavioral research, partly because their larger size makes certain surgical procedures and physiological measurements easier to perform.2PubMed Central. Serum Pharmacokinetics of a Highly Concentrated Buprenorphine Formulation in Older Female Sprague-Dawley Rats (Rattus norvegicus)
Behind the scenes, an entire infrastructure supports rodent research. Purpose-bred facilities maintain colonies of animals certified free of dozens of specific pathogens, including ectoparasites, bacteria, and viruses, to ensure that experimental results are not confounded by background infections.3SAGE Journals (Lab Anim). Moving from guideline recommendations to practical, reasonable proposals: Biosafety considerations, breeding performance and challenges during specific pathogen-free animal facility enlargement This level of standardization is harder and more expensive to achieve with larger animals, which is another reason rodents remain the default.
Zebrafish and Other Small Aquatic Models
If there is a rising star in the world of animal testing, it is the zebrafish. These small tropical fish have become one of the fastest-growing model organisms in biomedical research, and their use has expanded dramatically over the past two decades. Their appeal is partly practical: a single breeding pair can produce hundreds of embryos at a time, embryos develop outside the mother’s body, and their tiny size means large numbers can be housed in relatively small spaces.
But the real advantage is biological. Zebrafish embryos are transparent, which means researchers can watch organs form, blood vessels grow, and tumors develop in real time under a microscope. Their organ structures share meaningful similarity with human organs, and disease models can be established with relative ease.4PubMed Central. A high-throughput system for drug screening based on the movement analysis of zebrafish Zebrafish also show strong conservation of the molecular pathways involved in human disease, allowing researchers to do things like screen thousands of drug candidates quickly or validate genes suspected of playing a role in specific conditions.5PubMed Central. Recommendations of the Polish Zebrafish Society on the use of the zebrafish (Danio rerio) model in biomedical research
This combination of rapid development, genetic tractability, optical transparency, and high-throughput capacity has made zebrafish especially popular in early-stage drug screening and toxicology work. In many labs, they serve as a bridge between cell-culture experiments and mammalian studies: cheaper and faster than mice, but more biologically complex than cells in a dish.
Fruit Flies and Nematode Worms
Not all “animal testing” involves the kinds of animals people typically picture. Two of the most widely used organisms in genetics and developmental biology are the fruit fly (Drosophila melanogaster) and the nematode worm (Caenorhabditis elegans). Both have fully sequenced genomes, are easy to manipulate genetically, and reproduce fast enough that researchers can study multiple generations in weeks rather than years.6PubMed Central. Microfluidic tools for developmental studies of small model organisms — nematodes, fruit flies, and zebrafish
Fruit flies share a remarkable number of disease-related genes with humans. Researchers have used them to study everything from neurodegenerative diseases to the basic mechanisms of how embryos develop. C. elegans, which consists of fewer than a thousand cells and has a completely mapped nervous system, is a workhorse for understanding cell signaling, aging, and programmed cell death. Because these organisms are invertebrates, their use generally falls outside the regulatory frameworks that govern vertebrate animal testing in most countries, which means fewer ethical review hurdles and lower costs. That does not make them less scientifically valuable; it just means they occupy a different space in the research landscape.
Dogs in Safety Testing
The use of dogs in research is one of the most emotionally charged aspects of animal testing, and it is also one of the most misunderstood. Dogs are not used in the volumes that rodents are, but they play a specific and, under current regulations, often legally required role in pharmaceutical safety testing. Regulatory agencies in most countries require that new drugs be tested in at least two species before human trials begin: one rodent (usually a rat or mouse) and one non-rodent. The dog has long been the default non-rodent choice for many types of compounds.7PubMed. Use of the dog as non-rodent test species in the safety testing schedule associated with the registration of crop and plant protection products (pesticides): present status
Among dogs, beagles are overwhelmingly the breed used in research, selected for their relatively small size, docile temperament, and the existence of extensive historical data that makes it easier to compare results across studies. A review of published studies involving dogs found that beagles were the sole breed used in multiple experimental and clinical trial settings, with typical study sizes of around a dozen animals.8PubMed Central. What Kinds of Dogs Are Used in Clinical and Experimental Research? These are purpose-bred animals, raised in licensed facilities specifically for research rather than taken from shelters or the pet population.
The reasons dogs remain in this role are partly physiological and partly institutional. Their cardiovascular and gastrointestinal systems respond to drugs in ways that often predict human responses reasonably well for certain compound classes. But there is also a significant element of regulatory inertia: decades of dog toxicology data exist, and changing the default non-rodent species requires demonstrating that an alternative would be at least as predictive. Minipigs have been gaining ground as an alternative non-rodent species for some types of testing, though the shift has been gradual.
Rabbits and the Draize Test Legacy
Rabbits hold a distinctive and somewhat notorious place in the history of animal testing. The Draize test, developed in 1944, became the standard method for evaluating whether chemicals and cosmetic ingredients cause eye and skin irritation. For decades it was considered the gold standard, formally adopted as an international guideline. But the test involved applying substances directly to the eyes or shaved skin of rabbits and observing the damage over days, a procedure widely criticized as both cruel and scientifically imprecise.9PubMed Central. Alternatives to In Vivo Draize Rabbit Eye and Skin Irritation Tests with a Focus on 3D Reconstructed Human Cornea-Like Epithelium and Epidermis Models
Rabbit eyes differ from human eyes in several ways, including how quickly tears wash away irritants, which led to questions about how well results actually translated to human risk. This criticism, combined with growing public opposition, drove significant investment into alternatives. Today, reconstructed human tissue models made from cornea-like and skin-like cell layers can replicate much of what the Draize test measured, and many regulatory agencies now accept these in-vitro methods as partial or full replacements. The European Union banned animal testing for cosmetics ingredients entirely in 2013, which largely removed rabbits from that sector. However, rabbits are still used in some pharmaceutical and chemical safety studies, particularly where regulators have not yet validated an alternative for a specific endpoint.
Non-Human Primates
Non-human primates occupy the most contentious position in animal testing. Their close evolutionary relationship to humans makes them, in certain narrow applications, the only species that can provide relevant data. This is especially true for vaccines, biologics, and therapies that target immune pathways specific to primates, where rodent immune systems simply do not respond in comparable ways. Macaques (particularly rhesus and cynomolgus macaques) are the most commonly used primate species in research, followed by marmosets for some specialized applications.
Baboons, for instance, have been used in vaccine research because their immune responses closely mirror those of human infants, making them a rare model for testing vaccines intended for very young children. Studies have evaluated intranasal vaccines in infant baboons, using groups of around 20 animals, to demonstrate safety and protective efficacy before moving to human trials.10Biological Products. Prevention, Diagnosis, Treatment. Safety, immunogenicity, and protective efficacy of the intranasal live pertussis vaccine GamLPV in an infant monkey model Macaques have also been central to xenotransplantation research, serving as recipients in experimental transplants of genetically modified pig hearts to study rejection mechanisms and immunosuppression strategies.11PubMed. A Standardized Pig to Macaque Heterotopic Heart Xenotransplantation Model
The numbers involved are small compared to rodents. Primate studies are extraordinarily expensive, heavily regulated, and subject to extensive ethical review. In the United States, the National Institutes of Health retired most of its chimpanzees from research in 2015, and great apes are now essentially off-limits in most countries. The remaining primate research is concentrated in areas where no alternative species or method can provide equivalent data, such as neuroscience, infectious disease, and the development of complex biological therapies.
Chicken Embryos and Amphibians
Some widely used research animals rarely make headlines. The chicken embryo is one of them. Fertilized chicken eggs are cheap, readily available year-round, and the embryo develops rapidly inside a self-contained system that requires no maternal surgery or complex housing. The chorioallantoic membrane, a blood-vessel-rich layer just beneath the shell, provides a natural platform for delivering drugs or transplanting cells in a way that is essentially noninvasive.12PubMed Central. The Chicken Embryo: An Alternative Animal Model in Development, Disease and Pharmacological Treatment
Chicken embryos have contributed to fundamental discoveries in developmental biology, including processes common to all vertebrates. They have also been instrumental in understanding tumor growth and metastasis, studying blood vessel formation, assessing cancer drugs, and developing vaccines.13Developmental Biology. A scientific case for revisiting the embryonic chicken model in biomedical research In many jurisdictions, early-stage embryos are not classified as “animals” under welfare legislation, meaning their use may not require the same level of ethical approval as work with, say, a mouse. This regulatory distinction has made them attractive for screening studies that would otherwise demand larger numbers of mammals.
Amphibians fill yet another niche. The African clawed frog (Xenopus laevis) has been a model organism for decades, valued for its large, robust embryos and the fact that eggs can be obtained from adults year-round through in vitro fertilization. Researchers use Xenopus to study embryonic development, toxicology, neurobiology, endocrinology, and immune function. The frog’s genome shows a higher degree of similarity to mammalian genomes than many alternative invertebrate and vertebrate models, which gives its findings broader relevance.14PubMed Central. Xenopus laevis (Daudin, 1802) as a Model Organism for Bioscience: A Historic Review and Perspective With the recent development of gene-editing tools, Xenopus has also become an increasingly useful platform for genetics research.
How Regulations Shape Species Choice
It would be a mistake to think that researchers simply choose whatever animal they find most convenient. Regulatory requirements exert enormous influence over which species are used, particularly in pharmaceutical and chemical safety testing. As mentioned, most drug regulatory agencies require toxicity data from at least two species, a rodent and a non-rodent, before approving clinical trials in humans.7PubMed. Use of the dog as non-rodent test species in the safety testing schedule associated with the registration of crop and plant protection products (pesticides): present status The choice of which specific non-rodent to use depends on the compound: dogs for many small-molecule drugs, primates for biologics that only work in primate immune systems, and sometimes minipigs for dermal or oral formulations.
This two-species rule exists because no single animal perfectly predicts human biology. A side effect that shows up in rats might not appear in dogs, and vice versa. By requiring data from two evolutionarily distinct species, regulators are hedging their bets. The downside is that these requirements can be slow to change even when the science moves on. Agencies like the FDA and EMA have increasingly signaled openness to accepting data from organ-on-chip devices, computer models, and human-cell-based assays, but formal guidelines still lag behind the technology in many areas.
Academic research operates under a different framework. University-based studies are typically governed by institutional animal care committees that evaluate whether the proposed use of animals is justified, whether the number of animals is appropriate, and whether less sentient species or non-animal methods could achieve the same scientific goal. This is where the “3Rs” framework comes in: Replacement (using alternatives to animals when possible), Reduction (using the minimum number of animals necessary), and Refinement (minimizing pain and distress). The framework has been a guiding principle in animal research ethics since it was first proposed in 1959, though its application can be uneven. Pushing too aggressively on Reduction, for instance, risks using so few animals that the experiment produces inconclusive results, which wastes the animals that were used.15PubMed. Replacement, reduction and refinement
The Growing Role of Alternatives
The search for methods that can replace or supplement animal testing has intensified in the past decade. Organ-on-chip technology, which uses microengineered devices containing living human cells arranged to mimic the structure and function of real organs, is one of the most promising developments. These systems can simulate how drugs are absorbed, distributed, metabolized, and excreted by connecting multiple “organ” chips together, creating something like a miniature human body on a laboratory bench. Researchers see multi-organ-on-chip platforms as a way to get better predictions of how drugs will behave in people, potentially replacing some of the animal studies currently required.16Current Opinion in Toxicology. The potential of multi-organ-on-chip models for assessment of drug disposition as alternative to animal testing
Other alternatives include computational models that predict toxicity based on a chemical’s structure, 3D cell cultures that better replicate human tissue than traditional flat cell layers, and organoids, which are miniature self-organizing organ-like structures grown from stem cells. None of these has fully replaced animal testing for any major regulatory category yet, but each is chipping away at specific use cases. The reconstructed human tissue models that have partially replaced the Draize rabbit eye test are a concrete example of this progress in action.
In the United States, the FDA Modernization Act 2.0, signed in late 2022, formally removed the longstanding requirement that drugs must be tested on animals before human trials, opening the door for companies to use alternative methods if they can demonstrate equivalence. This does not mean animal testing has ended or will end soon; rather, it means the legal framework no longer mandates it as the only path. Companies still have to convince regulators that their safety data, whatever its source, is robust enough to justify putting a drug into people. For the foreseeable future, that will involve animals for most complex therapies, but the trajectory is clearly toward using fewer of them and using them more selectively.
Pigs as Translational Models
Pigs deserve their own mention because their role in research is growing and is distinct from the traditional laboratory species. Their cardiovascular systems, skin, and digestive tracts bear a closer resemblance to human anatomy than those of most rodents or dogs. Minipigs have become an accepted alternative to dogs for some pharmaceutical toxicology studies, and standard-sized pigs are used in surgical training, wound-healing research, and increasingly in xenotransplantation, the transplantation of organs across species.
Genetically modified pigs, with certain genes knocked out to reduce immune rejection, are being used in experimental organ transplants into primates as a step toward eventually using pig organs in human patients.11PubMed. A Standardized Pig to Macaque Heterotopic Heart Xenotransplantation Model This line of research has already produced headline-grabbing cases of pig hearts and kidneys being transplanted into human patients, though these remain experimental. The pig’s dual role as both a testing model and a potential source of transplantable organs makes it unique among laboratory animals.
What the Numbers Look Like by Species
Exact global figures are hard to pin down because countries count differently, some exclude certain species, and many nations do not report at all. In the European Union, which has the most detailed reporting requirements, mice consistently account for more than half of all procedures on animals, followed by rats, fish, and birds. Dogs, cats, and primates together make up less than one percent of the total but receive outsized public attention because of their status as companion animals or close evolutionary relatives.
In the United States, the Animal Welfare Act requires annual reporting of animals used in research, but it explicitly excludes rats, mice, and birds bred for research from its counts. This means the official USDA numbers dramatically undercount total animal use. Independent estimates, which attempt to include rodents and fish, suggest that the true number of animals used annually in U.S. research is in the tens of millions, with mice and rats making up the overwhelming majority.
Zebrafish and invertebrates further complicate the picture. Many jurisdictions do not count early-life-stage fish or invertebrates like fruit flies and nematodes as “animals” for regulatory purposes, so these organisms often go entirely uncounted despite being used in enormous numbers. The trend in regulatory reporting has been toward more inclusive counting, but meaningful global totals remain elusive.
Why Certain Species Persist Despite Alternatives
A common question is why animals are still used at all when so many alternatives exist. The honest answer is that no alternative yet replicates the full complexity of a living organism. A drug can look perfectly safe in cell cultures and computer models but still cause liver damage in a living animal because of an unexpected interaction between the drug’s breakdown products and the immune system, the gut microbiome, or the hormonal environment. Whole-animal testing, for all its limitations, captures these systemic interactions in a way that isolated systems currently cannot.
Species-specific persistence also comes down to historical data. Decades of dog toxicology studies have created a reference library that allows researchers to compare a new drug’s effects against thousands of previous compounds tested in the same species. Switching to a different species means losing that comparative context. The same logic applies to specific strains of mice: a researcher studying Alzheimer’s disease in a particular transgenic mouse strain can draw on hundreds of published studies using that same strain, making results interpretable in a way that starting fresh with a new model would not be.
Regulatory agencies, too, are understandably conservative. Their mandate is to protect public safety, and accepting unfamiliar methods means accepting unfamiliar risks. The shift away from animal testing is happening, but it is happening through slow accumulation of validation data rather than through any single breakthrough. Each time a non-animal method is formally validated for a specific regulatory endpoint, the number of animals needed drops a little. Over time, these incremental changes add up, but the process is measured in decades rather than years.