On September 21, 2026, the U.S. Food and Drug Administration (FDA) issued a rule that, on paper, only changes vocabulary. It replaces “animal tests” and “animal studies” with “nonclinical tests” and “nonclinical studies” throughout its rules for drugs and biologics. The FDA says the rule adds no new requirements or costs. Both statements are true. They are also beside the point. Regulatory lawyers, FDA reviewers, and procurement committees read regulations, and the words they read often shape what they are willing to pay for. This piece walks through what the rule does, and its impact from four seats at the table: the drug companies that must prove their products are safe, the companies selling animal-free testing tools (NAMs), the contract research firms that run most safety studies, and the companies that breed and supply research animals.
What the rule says
The rule amends five parts of Title 21 of the Code of Federal Regulations (CFR), the rulebook the FDA uses to regulate food, drugs and medical products. The parts affected (312, 314, 315, 361 and 601) cover the application to start human trials, new drug approvals, biologic licenses, and two narrower areas dealing with radioactive drugs used in diagnosis and research.
It also adds a definition. A “nonclinical test” can now be done in a dish of cells (in vitro), on a computer (in silico), with chemical reactions alone (in chemico), or in a living nonhuman animal (nonhuman in vivo). That tracks a 2022 federal law, the FDA Modernization Act 2.0, passed alongside the Food and Drug Omnibus Reform Act, which listed cell-based assays, organ chips, computer models and bioprinting as acceptable ways to show a drug is safe enough to try in people. The industry’s umbrella term for these animal-free tools is new approach methodologies, or NAMs.
Three details matter for timing and scope:
- Timing. The rule takes effect February 4, 2027, unless the FDA receives “significant adverse comments” by December 7, 2026. If it does, the agency withdraws the rule and moves forward with a companion proposed rule that says the same thing. Objections can delay the change; they are unlikely to kill it.
- A precedent library. The FDA also published a database of 25 real examples in which non-animal data was accepted in past drug reviews. For cautious companies, precedent is often more persuasive than policy.
- What it does not do. Animal testing is not banned. Evidence standards do not change. Also, medical device rules are not among the amended parts, so makers of 3D-printed implants get nothing from this rule directly.
Why vocabulary matters
The FDA issued this as a “direct final rule,” a fast-track process reserved for changes the agency considers noncontroversial. To qualify, the FDA said it does not expect the rule to affect industry practice. That is the legal framing. The commercial reality is different.
Under the old wording, an animal study was the default and anything else was an exception a company had to argue for. Under the new wording, animal and non-animal methods sit in the same sentence as equal options. The burden of proof has not moved: a company must still show that its chosen method is reliable for the specific question it is answering, which regulators call “context of use.” What has moved is the starting point of the conversation.
The rule also caps a string of related steps. In April 2025, the FDA announced a plan to phase out animal testing requirements for monoclonal antibodies, a large class of lab-made antibody drugs. It followed with draft guidance in December 2025 on reducing monkey studies for those drugs and broader draft guidance on NAMs in March 2026. Taken together, the direction is clear.
Pharmaceutical companies: more options, same homework
The case for moving. The FDA’s own framing is that more than 90% of drugs that look safe and effective in animals go on to fail in human trials. Every one of those failures is a sunk cost. A human-cell model that catches a liver problem before a drug reaches people saves money and time. The most cited evidence remains the Emulate Liver-Chip study: across 27 drugs tested blind, the chip correctly flagged 87% of drugs known to injure the liver and raised no false alarms on the safe ones (Ewart et al., Communications Medicine, 2022).
Biology adds a second argument. Many modern drugs, including antibody drugs and gene therapies, are built to act on human-specific targets. Often the only animal that responds is a monkey, and monkeys are expensive, slow to source and in short supply. For those drugs, a human-based model is not just an ethical preference; it may be the more relevant test.
Where do NAMs pay off first? At a recent 3DHEALS panel on NAMs, Graham Craig of VoxCell named four decisions they can inform earlier than clinical trials: whether a drug engages its target in human tissue, whether it has the intended effect, what side effects it may cause, and whether it reaches the right tissue. Framing a model around one of those questions is easier to defend than claiming it captures all of human biology.
The case for caution. Regulatory teams are paid to avoid delay, and the costliest delay is an FDA order that pauses a human trial. No one has lost their job for running the standard rat study. Drug companies also file in Europe, Japan and elsewhere, and international safety guidelines still expect animal data for many endpoints. So, in the near term, most companies will run NAMs alongside animal studies rather than instead of them. That raises costs before it lowers them.
Bottom line. Drug companies gain an option, not an obligation. Expect early adoption where animal models are weakest (human-specific biologics, liver and heart safety screens) and where NAMs are used early to weed out bad compounds before the formal safety package. Expect animal studies to stay in the final submission for years.
NAMs providers: from pilot projects to purchase orders
The case for optimism. Makers of organ chips, organoids (tiny lab-grown organ-like tissues), bioprinted tissue and computer models have spent a decade selling into research budgets, not regulatory budgets. Every sale came with the same question: will the FDA accept this? The rule does not answer that question for any specific product, but it removes the hardest version of it. The category is now written into the regulation, and the 25-case database gives sales teams real examples to point to. Clearer rules also lower perceived risk for investors, which should help fund the unglamorous work of scaling up manufacturing and automation.
The case for realism. The bar is still validation, and the first 25 examples mostly involve standardized skin and eye irritation tests that took years to validate, not organ chips or bioprinted tissue. The Emulate study is one organ, one type of harm and a company-led study. Buyers will now ask harder questions: Does the model give the same answer in different labs? Does one batch behave like the next? How many compounds can it test per week, and at what cost per data point? Providers that cannot answer with data will find that the rule changed the conversation but not the purchase decision.
The view from founders. On Episode 122 of The Lattice Podcast, Lowry Curley, PhD, founder of Luna LifeSci and of 28bio (formerly AxoSim), argues the FDA is at an inflection point where organ chips and lab-grown mini-brains move from niche to routine. He notes that failure rates for drugs that pass animal testing climb to about 94% in neurology, and he frames the open question for investors as who will consolidate the field. On the same 3DHEALS panel, Mike Clements of Axion BioSystems cautioned that flexibility cuts both ways: the scientific bar has not changed, and more choice of methods can mean more variability unless the industry standardizes how models are built and checked.
Bottom line. The rule is a marketing asset. Revenue will follow validation data, not vocabulary. Providers that publish head-to-head comparisons against known drugs will pull ahead.
Contract research organizations: the integrators in the middle
The case for gaining. Contract research organizations (CROs) run most of the outsourced safety testing in drug development. Drug companies generally buy a finished safety package, not individual technologies, and that plays to the CROs’ strength. A large CRO can bundle computer models, cell-based tests and animal studies into a single program and tell the client which mix a reviewer is likely to accept. Several of the largest CROs have publicly invested in alternative methods for exactly this reason.
CROs also hold something NAMs start-ups lack, which is decades of animal data on known drugs. That history is what a new method must be measured against, which makes CROs natural partners for validation work. And during the transition, validation means running the new test and the old test side by side. That is more work, and CROs get paid for work.
The case for strain. Animal facilities are expensive fixed assets. If study volumes decline, the cost of each remaining study rises. Smaller CROs that only run animal studies and lack the capital to build cell-culture or computing teams, face slow margin erosion as early screening work moves to faster, cheaper platforms.
Bottom line. Large CROs are likely to win the transition, provided they manage the gradual shrinking of their animal capacity. The squeeze falls on smaller, animal-only labs.
Animal model providers: less volume, more specialization
The case for resilience. Nothing is banned. Many questions remain hard to answer outside a living body: how a drug spreads through and leaves the body over time, effects on fertility and pregnancy, cancer risk from long-term exposure, and interactions between organs. Demand may even grow for specialized animals, such as genetically engineered or “humanized” mice carrying human genes or cells, because they answer questions that chips cannot yet handle.
The case for concern. The FDA’s recent guidance targets monkey studies for antibody drugs first. Monkeys are the most expensive research animal, so any drop in that segment hits revenue harder than a similar drop in rodents. Routine rodent studies used for early screening are also exposed, as that work is the easiest to move onto cell-based and computer tools. Animal suppliers could file objections before December 7, but because the backup proposed rule is identical, objections would mostly buy time, and the public relations cost of opposing an “alternatives to animal testing” rule is high.
Bottom line. Expect a slow erosion rather than a cliff, with monkeys most exposed. The providers best placed to adapt are those moving toward well-characterized, specialized models and partnering with NAMs companies to offer combined evidence packages.
What it means for 3D bioprinting
Bioprinting is named in the 2022 law, but it does not appear among the first 25 examples in the FDA database. Today it is used mostly inside companies for early drug discovery, not in the safety packages submitted to regulators. The obstacles are practical: keeping printed cells alive and functioning over time, making each batch behave like the last, and producing tissue at a scale and price that drug companies will buy. Reviewers currently trust standardized organ chips and established tissue models more than custom-printed tissue.
Early examples show what the path looks like. At the 3DHEALS NAMs event, FluidForm Bio described bioprinted human heart tissue that beats, can be pushed into an irregular rhythm, and returns to normal with a standard drug (lidocaine). Pranav Joshi of Bioprinting Laboratories argued that the real burden for customers is preparation, and that quality checks with clear pass/fail criteria need to run through the whole life of an organoid, not just at the end.
For a bioprinted model to reach a regulatory filing, it will need a clearly defined question it answers, results that repeat across batches and labs, benchmarking against drugs with known effects, and a data package a reviewer can easily follow.
What to watch
- December 7, 2026: Whether the FDA receives significant objections, which would delay but likely not stop the change.
- February 4, 2027: The rule’s scheduled effective date.
- Growth of the use-case database: Whether organ-chip or bioprinted examples join the skin and eye tests.
- Final guidance: How much detail the final versions of the 2025 and 2026 draft guidances give on validation.
- Devices: Whether the FDA extends similar language to medical device rules, which this rule leaves untouched.
References
- Food and Drug Administration. “Nonclinical Testing Terminology” (direct final rule). Federal Register, 91 FR 59988, September 22, 2026. https://www.federalregister.gov/documents/2026/09/22/2026-19350/nonclinical-testing-terminology
- Food and Drug Administration. “FDA Updates Regulations to Advance Innovative Alternatives to Animal Testing” (press release). September 21, 2026. https://www.fda.gov/news-events/press-announcements/fda-updates-regulations-advance-innovative-alternatives-animal-testing
- Food and Drug Administration. “New Approach Methodologies (NAMs).” https://www.fda.gov/science-research/science-and-research-special-topics/new-approach-methodologies-nams
- Food and Drug Administration, Center for Drug Evaluation and Research. “New Approach Methodologies (NAMs) Database: Use Case Examples.” https://www.fda.gov/drugs/cder-streamlined-nonclinical-studies-and-acceptable-new-approach-methodologies-nams/new-approach-methodologies-nams-database-use-case-examples
- Food and Drug Administration. “FDA Announces Plan to Phase Out Animal Testing Requirement for Monoclonal Antibodies and Other Drugs” (press release). April 10, 2025. https://www.fda.gov/news-events/press-announcements/fda-announces-plan-phase-out-animal-testing-requirement-monoclonal-antibodies-and-other-drugs
- U.S. Congress. S.5002, FDA Modernization Act 2.0, 117th Congress (2022). https://www.congress.gov/bill/117th-congress/senate-bill/5002
- Regulatory Affairs Professionals Society (RAPS). “FDA to remove ‘animal testing’ terminology from its regulations.” September 2026. https://www.raps.org/resource/fda-to-remove-animal-testing-terminology-from-its-regulations.html
- Pharmaceutical Technology. “The FDA Finalizes Rule Recognizing Non-Animal Testing Methods, Building on Year-Long Reduction Push.” September 2026. https://www.pharmtech.com/view/the-fda-finalizes-rule-recognizing-non-animal-testing-methods-building-on-year-long-reduction-push
- BioPharm International. “FDA Formalizes Non-Animal Testing Methods in New Rule.” September 2026. https://www.biopharminternational.com/view/fda-formalizes-non-animal-testing-methods-new-rule
- AllSci. “FDA modernizes drug regulations to formalize support for NAMs.” September 2026. https://allsci.com/news/regulatory-policy/fda-nonclinical-testing-terminology-issues-direct/
- BioBuzz. “FDA Updates Regulations to Advance Innovative Alternatives to Animal Testing.” September 2026. https://biobuzz.io/news/fda-updates-regulations-to-advance-innovative-alternatives-to-animal-testing/
- Ewart L, et al. “Performance assessment and economic analysis of a human Liver-Chip for predictive toxicology.” Communications Medicine 2022;2:154. doi:10.1038/s43856-022-00209-1. https://doi.org/10.1038/s43856-022-00209-1
- 3DHEALS. “Episode 122: New Approach Methodologies (NAMs) with Dr. Lowry Curley.” The Lattice Podcast, August 8, 2026. https://3dheals.com/episode-122-new-approach-methodologies-nams-with-dr-lowry-curley/
- 3DHEALS. “Event Recap: New Approach Methodologies (NAMs).” August 26, 2026. https://3dheals.com/event-recap-new-approach-methodologies-nams/
Glossary
- Biologic: A medicine made from living cells, such as an antibody, vaccine or gene therapy, rather than from chemical synthesis.
- Bioprinting: Using a 3D printer to lay down living cells and support materials in layers to build tissue-like structures.
- Clinical hold: An FDA order that delays or pauses a human trial until safety concerns are resolved.
- Code of Federal Regulations (CFR): The official collection of U.S. federal rules. Title 21 covers food, drugs and medical products.
- Context of use: The specific question a test is meant to answer, and the conditions under which its results can be trusted.
- Contract research organization (CRO): A company that runs laboratory studies or clinical trials on behalf of drug developers.
- Direct final rule: A fast-track rule for changes an agency expects to be noncontroversial. It takes effect unless significant objections arrive.
- In vitro, in silico, in chemico, in vivo: Tests done in cells or tissue outside the body, on a computer, through chemical reactions alone, and in a living organism.
- Monoclonal antibody: A lab-made antibody designed to lock onto one specific target in the body.
- New approach methodologies (NAMs): Testing methods that do not rely on animals, including cell-based tests, organ chips, organoids and computer models.
- Nonclinical test: Any safety or effectiveness test done before or outside human trials, with or without animals.
- Organ chip: A small device lined with human cells that mimics how an organ works, including fluid flow and tissue interaction.
- Organoid: A tiny, self-organizing cluster of lab-grown cells that resembles part of an organ.
- Validation: Evidence that a test gives accurate, repeatable results for its intended context of use.
Comments