In vitro testing is any experiment performed on biological material outside a living organism, typically in a lab dish, test tube, or controlled artificial environment. The Latin phrase “in vitro” literally means “in glass,” a nod to the glass vessels scientists originally used. This contrasts with “in vivo” testing, which takes place inside a living animal or human. In vitro methods are used across medicine, from screening new drugs for safety to running the blood tests your doctor orders at a routine checkup.
How In Vitro Testing Works
The core idea is isolation. Scientists take cells, tissues, proteins, or other biological components out of the body and study them in a controlled setting. This lets researchers observe specific reactions without the complexity of an entire living system interfering. A liver cell in a dish, for example, can be exposed to a new chemical compound so researchers can watch exactly how it responds, something that would be impossible to observe directly inside a person.
The simplest and most common setup is two-dimensional cell culture: cells grown as a single flat layer on a plastic surface. Most cell-based lab work still uses this approach because it’s inexpensive, well-established, and easy to scale up. But flat layers of cells don’t behave quite like cells in the body, which exist in three-dimensional structures surrounded by other cell types. To get closer to real biology, researchers now also grow cells into 3D clusters called spheroids, either within a supportive scaffold material or freely suspended in liquid using techniques like hanging drop methods. These 3D cultures better mimic how tumors grow, how tissues heal, and how organs function.
Drug Development and Safety Screening
In vitro testing plays a critical role in getting new medicines from the lab to the pharmacy shelf. Before a drug candidate ever reaches a human clinical trial, it goes through rounds of in vitro screening to answer basic safety questions. Can the compound cross cell membranes? How quickly does the body break it down? Does it damage liver cells or cause DNA mutations?
These early screens evaluate what pharmacologists call ADME properties: how a drug is absorbed, distributed through the body, metabolized, and eliminated. Before these in vitro ADME screens became standard practice, roughly 40% of drug candidates failed in clinical trials because of poor absorption or metabolism. After pharmaceutical companies adopted in vitro screening methods, that failure rate dropped to less than 10% by the year 2000.
Liver toxicity is one of the biggest reasons drugs get pulled from the market, so specialized in vitro liver models get particular attention. Researchers use human liver cells grown in layered “sandwich” cultures that maintain the cells’ natural structure, allowing them to detect specific types of liver damage that simpler cultures would miss. When thousands of candidate molecules emerge from early discovery, high-content screening systems can test them rapidly, flagging the most dangerous compounds before significant money is spent on further development.
One study evaluating an in vitro cytotoxicity assay in breast cancer found it predicted actual tumor response to chemotherapy drugs with 89% sensitivity, meaning it correctly identified effective drugs nearly nine times out of ten. The specificity was lower at 53%, so it was better at catching what works than ruling out what doesn’t. Still, that kind of predictive power early in the process saves enormous time and resources.
In Vitro Diagnostics in Clinical Medicine
If you’ve ever had blood drawn for a cholesterol panel, taken a COVID test, or received genetic screening results, you’ve used in vitro diagnostics. The FDA defines these as tests performed on samples like blood or tissue that have been removed from the body. They’re one of the most common forms of in vitro testing, and they touch nearly every area of medicine.
The category is broad. It includes simple urine test strips, complex immune-based assays that detect specific antibodies or proteins, and molecular tests like PCR that amplify tiny traces of genetic material to identify infections. More recently, next-generation sequencing tests scan a person’s entire DNA to detect genomic variations linked to cancer risk, inherited conditions, or drug sensitivities. All of these qualify as in vitro diagnostics because the analysis happens outside the body, on a sample taken from it.
How It Compares to Animal Testing
In vitro methods offer clear advantages over in vivo animal studies, but they also have real limitations. The biggest benefit is ethical: in vitro testing reduces the number of animals used in research and avoids procedures that can cause significant pain or distress. It also allows researchers to work directly with human cells, which sometimes behave differently from animal cells in ways that matter for predicting human drug responses.
The tradeoff is complexity. A dish of liver cells can tell you whether a compound is toxic to those specific cells, but it can’t tell you what happens when that compound interacts with the immune system, the kidneys, and the gut simultaneously. Living organisms are interconnected systems, and in vitro models capture only a piece of that picture. There are also practical concerns: antibodies produced in vitro sometimes lose their proper structure compared to those produced in animals, potentially affecting how they function in later experiments. And for small-scale production of certain biological products, in vitro methods can cost anywhere from half to six times as much as animal-based methods, depending on the equipment and purification steps required.
Organ-on-a-Chip Technology
One of the most significant recent advances in bridging the gap between simple cell cultures and whole-body complexity is the organ-on-a-chip. These are small devices, often no bigger than a USB drive, that contain tiny channels lined with living human cells. Fluid flows through the channels to simulate blood flow, and the chips can replicate mechanical forces like the stretching of lung tissue during breathing.
Researchers have built chips that model a wide range of organs and diseases. Lung chips can simulate nanoparticle toxicity, viral infections, and the scarring seen in fibrotic disease. Liver chips test drug toxicity and liver-specific functions. Gut chips replicate intestinal barrier function, drug absorption, and inflammatory reactions. Brain barrier chips screen drug permeability into the central nervous system. Skin chips can test cosmetic and pharmaceutical irritants, producing results that meet international safety standards.
What makes these chips valuable is that they capture interactions between different cell types under conditions that resemble the body’s actual mechanical and chemical environment. A lung chip doesn’t just contain lung cells; it recreates the air-blood barrier where gas exchange happens, something a flat culture dish cannot do.
The Regulatory Shift Away From Animal Testing
In late 2022, the U.S. Congress passed the FDA Modernization Act 2.0, a landmark law that explicitly authorized the use of non-animal alternatives, including cell-based assays and computer models, to support applications for new drugs. The law removed a longstanding requirement to use animal studies for certain types of biological drug applications.
The FDA has published a roadmap aiming to make animal studies “the exception rather than the norm” for preclinical safety testing within three to five years. In the near term, the agency is encouraging drug companies to submit data from organ chips, cell-based models, and computer simulations alongside traditional animal data to build a track record. A central database of validated non-animal methods, called CAMERA, launched a beta version in mid-2025. The FDA has also signaled it will offer regulatory relief, such as requiring fewer animal study replicates, to companies that provide strong non-animal data.
Machine Learning and In Vitro Data
Artificial intelligence is increasingly paired with in vitro experiments to extract more information from lab data. In one recent approach, researchers collected published in vitro drug release data covering variables like acidity, particle size, drug solubility, and molecular weight, then fed it into machine learning algorithms. The models identified patterns that would be difficult for humans to spot: acidic or alkaline environments accelerate drug release, higher solubility increases release rates, and smaller particles release their contents more efficiently.
The key validation step was that researchers then designed new in vitro experiments based on the machine learning predictions and found the lab results matched. This kind of feedback loop, where computational models guide physical experiments and experiments refine the models, is becoming a standard approach for analyzing the large, complex datasets that modern in vitro screening generates.