A mouse is held against a flat surface. Someone presses a thumb, finger, pen, pair of closed scissors or metal rod against the back of her neck. Her body is pulled in the opposite direction.
The intention is to separate the vertebrae at the top of her spine, damage her spinal cord and disrupt the blood supply to her brain. When performed correctly, cervical dislocation is supposed to cause rapid unconsciousness and death. The problem is that British laboratories do not appear to agree on how to perform it correctly.
A study of 317 people working with mice and rats across UK research institutions found widespread differences in the techniques, physical actions and tools used to kill them. There is no standardised method, no scientifically validated tool designed specifically for the procedure and little agreement about what a successful dislocation even requires.
Cervical dislocation is reportedly the preferred killing method of 57.5% of UK laboratory personnel.
No Standard Method
Mice and rats account for around 70% of all experimental procedures involving mammals in the UK. When experiments end, when animals are no longer wanted for breeding, or when they have been produced in excess, they are killed.
UK law permits several methods, including anaesthetic overdose, carbon dioxide exposure, concussion and cervical dislocation. The law describes these methods as acceptable. That does not mean they are painless, reliable or ethically defensible.
For mice, 69.4% of survey respondents reported that cervical dislocation was usually performed manually, without equipment. Around a quarter used a mechanical aid. For rats, the pattern was reversed. More than two thirds used some form of tool, while only 15% relied primarily on their hands.
What tools? Metal rods. Closed scissors. Pens. Cage scrapers. Rulers. Broom handles. Spanners.
Ordinary objects that were never designed, tested or validated for killing animals. The researchers warned that pens may break, scissors may slip and blunt objects may crush the neck rather than separate the vertebrae. An inaccurate or incomplete dislocation could delay unconsciousness, cause asphyxiation or prolong the animal’s death. This is what can hide behind the word “euthanasia.” A mouse pinned beneath a pen. A rat killed with a cage scraper.
Laboratories Do Not Even Agree On The Actions
For cervical dislocation to cause rapid unconsciousness, the separation should occur as close to the skull as possible. But survey respondents did not agree on which actions were necessary to achieve this. Some emphasised pushing down. Others pushed forwards. Some pulled the tail horizontally or upwards. Some stretched the neck. Others twisted or rolled the animal’s body. Evidence from research involving chickens suggests that a rapid stretch and twist together cause more complete separation. Applying only one part of the movement can result in less damage to the spinal cord, a slower loss of consciousness and greater suffering. More than half of those using mechanical methods on mice and rats instead emphasised a forceful downward push. The study’s authors warned that this may be closer to crushing the neck, potentially causing a slower death through asphyxiation.
Even the rules within individual institutions were unclear. More than one in five respondents working with mice did not know whether their laboratory had a formal procedure. For rats, 27.5% said no standard operating procedure existed at all. The researchers concluded that practice appeared to be driven by local processes and personal preferences rather than empirical evidence. Despite this, most respondents still described the method as “humane.”
That confidence is revealing.
When a killing method is familiar, legal and performed inside a laboratory, the label often becomes more important than what happens to the animal.
The Animals Who Never Reach An Experiment
The violence of animal research does not begin with an injection or end when an experiment is completed. Many animals are bred for research and killed without ever being included in a study.
In Switzerland, around 1.2 million animals are bred for scientific purposes each year, but only roughly half are used directly in experiments. The remaining 600,000 may have the wrong sex, age or genetic characteristics, or simply exceed the number researchers require.
They are classified as surplus.
That word erases an entire life.
These animals are confined under laboratory conditions and killed because the system has no further use for them. Their deaths may not appear in official figures for animals “used” in experiments, even though they were deliberately brought into existence by the research industry.
Their confinement causes harm too. Mice and rats have strong drives to burrow, forage, explore and form complex social relationships. Standard laboratory cages restrict these behaviours. Research has associated such environments with boredom, frustration, anxiety, stress and depression, as well as increased rates of some physical illnesses and premature death. But many legal systems focus narrowly on what happens during an authorised experimental procedure. Confinement, killing and the loss of an animal’s future may be excluded from the formal calculation entirely.
Better Science Already Exists
Defenders of animal experimentation often present the choice as simple: harm animals or abandon medical progress. That choice is false.
Researchers are already developing methods that are faster, cheaper and more relevant to human biology.
In Switzerland, scientist Jimeng Wu has created an AI-powered “virtual mouse” for nanomedicine research. The model draws on data from 18 existing studies to predict how nanoparticles will behave inside the body, allowing researchers to screen potential interventions before manufacturing them or testing them on living animals.
Researchers in Germany have developed genESOM, a generative AI system trained on small experimental datasets. It produces synthetic data designed to reflect genuine laboratory results and could reduce animal use in some preclinical research by between 30% and 50%.
Other methods include human cell cultures, organoids, organs-on-chips, computer modelling and 3D bioprinted human tissues.
These are not speculative technologies waiting for some distant scientific breakthrough.
An analysis of biomedical and toxicology papers published between 2003 and 2022 found that research using only non-animal methods had outnumbered animal-based research throughout the entire period. By 2022, 73% of the papers studied used non-animal methods alone. Another 24% combined animal and non-animal methods. Just 3% relied solely on animals. In the UK, researchers published more than four non-animal-only papers for every animal-based paper.
Animal experimentation is not the uncontested foundation of modern science. In several major fields, researchers are already moving beyond it.
Tradition Disguised As Necessity
So why does animal experimentation continue?
Part of the answer lies in scientific culture.
A survey of researchers in India found that 56% had been asked by peer reviewers to add animal experiments to studies that originally contained none. These demands contributed to rejected papers, withdrawn submissions, publication in lower-impact journals and reduced funding opportunities. Around one in five respondents admitted conducting animal experiments purely because they expected reviewers to demand them later.
Animals were not necessarily used because the research required them. They were used because the publishing system expected their bodies to appear in the evidence.
Scientific authority can also weaken moral resistance.
In one experiment, participants were asked to administer increasingly toxic doses to what they believed was a living fish. More than half continued until they believed they had killed the fish. Those encouraged to think positively about science were more likely to continue despite signs of distress.
Across a separate sample of more than 31,000 Europeans, stronger pro-science attitudes predicted greater support for animal experimentation.
This does not mean science is the enemy. It means scientific authority must not become a moral exemption.
A laboratory coat does not turn violence into neutrality. A research licence does not make an animal disposable. The possibility of human benefit does not erase the individual whose body is being used to pursue it.
Replace, Do Not Refine
The researchers behind the UK cervical dislocation study call for validated tools, clearer procedures, stronger training and more consistent regulation.
Those changes might reduce some immediate suffering.
But a purpose-built device would still be designed to break an animal’s neck. A perfectly written procedure would still end with a dead mouse. A more reliable killing method would not explain why she was bred, confined, experimented upon and discarded in the first place.
The real scandal is not simply that laboratories are killing mice and rats with improvised objects.
It is that millions of sensitive, social animals are treated as disposable research materials, while institutions debate how to kill them more consistently.
Science is already capable of moving beyond animal experimentation. AI, human tissues, organs-on-chips and computer models are not concessions to ethics at the expense of progress. They are part of better, more human-relevant research.
The future of science should not depend on finding a more efficient way to break an animal’s neck. It should depend on no longer doing it at all.
The question is not, Can they reason?, nor Can they talk? but, Can they suffer? — Jeremy Bentham