Monday is shipping day at MatTek. A truck pulls up to its red brick lab outside of Boston to load box after box, all kept at a cool 39 degrees. The precious, perishable cargo is human skin---thousands of dime-sized pieces in plastic dishes that add up, altogether, to about two whole adult humans’ worth. Every week.

It’s not harvested from people, though. MatTek grows its own human skin, and then sells it to companies that want it---companies that make laundry detergent, makeup, toilet bowl cleaner, anti-aging creams, tanning lotion.

Without lab-grown skin, these companies would be testing products on animals, usually rabbits, shaved to expose patches of naked skin. This practice is straight-up illegal for cosmetics in Europe now, and increasingly ethically dubious everywhere else. With animal testing on the wane, MatTek---along with its chief competitor, Episkin, a subsidiary of L’Oreal---have become go-to sources for synthetic human skin.

Testing on lab-grown medallions isn’t just an ethical choice: It’s practical. “They are a much better simulation of human skin than animals are,” says Carol Treasure, whose company XCellR8 test products for brands like Lush Cosmetics and uses MatTek’s products for some of the work. But truth be told, you wouldn’t recognize MatTek’s skin as human, or even as skin. To the naked eye, the skin---less than fifth of a millimeter thick---looks like thin circles of clear jello.

So, too, do MatTek’s other tissue models: bits of eye, lung, intestinal, vaginal, and mouth tissue, all grown for testing. The company’s process reduces body parts to their most essential cells---turning surgical waste from biopsies, tummy tucks, and circumcisions into a reliable and standardized product lines, all of which turn into those translucent discs. “You wouldn’t be able to tell the difference just from looking at them,” says president Mitch Klausner.

That uniformity is MatTek’s greatest challenge---and advantage. Human skin is hugely diverse. Take any two people of the same age, sex, and race, and one might be oilier or more sensitive or drier than the other. MatTek’s skin tissue model must react the same way to the same chemicals year after year, even if the originals cells came from two different people and two different parts of the body. It runs a highly tuned skin factory.

To start, MatTek needs a small but steady supply of real human skin, which they could use as seed material to grow large quantities in the lab. Most human cells can only replicate so many times before they die. And thus we all age and we perish.

The company has forged partnerships with local hospitals to get surgical waste from cosmetic surgeries and circumcisions, where patients---or their parents---have agreed to donate excess skin to research. As MatTek has expanded other tissues, it also worked with the National Disease Research Interchange, an organization funded by the National Institutes of Health to disseminate tissues from deceased donors for research.

Klausner was not keen to go into the details of acquisition, citing the anonymity of donors. Indeed, MatTek knows very little about the patients from whom it gets donations. And patients themselves might have a hard time tracing their own donated tissues. When I asked if anyone has ever attempted to find what happened to the skin from their surgery, Klausner dismissed the very idea.

Patients may not even know where to look given that consent forms---part of the standard hospital admissions paperwork---don’t usually specify how tissue will be used. It could go to a patient’s doctor or to a researcher at another university or to a company. “There’s a lot about how research is conducted in the US that the public doesn’t understand,” says Michelle Lewis, a bioethicist at Johns Hopkins Berman Institute of Bioethics.

The process is extremely opaque. But if you’ve had a circumcision or tummy tuck or breast surgery, especially in a Boston-area hospital, your cells could have made it to MatTek, been expanded to cover two football fields worth of skin, and been sent all over the world to labs that test chemicals and skin creams and drugs.

MatTek does keep track of some identifying factors, though. It sorts its skin tissues by age, sex, and race, depending on the intended test. Some are harder to source than others. The company makes some skin containing pigmentation cells---to test products like tanning lotion or lightening creams---and the Asian market is big, especially for skin lightening creams. “Asian skin is a little harder to get,” though, says Klausner: Asian parents are less likely to circumcise their baby sons.

To keep its product mass-market consistent, MatTek's technicians first use enzymes to break down that original piece of donated skin into individual cells. Epidermis actually contains many different types of cells, but the main ones are keratinocytes. So the technicians take the keratinocytes and grow them in a single layer in petri dishes. Then they isolate individual keratinocytes and use them to seed porous inserts in plastic wells.

You could stop here, but MatTek doesn’t just want to grow a mass of undifferentiated skin cells in a dish. It wants to grow skin tissue---with layers of cells that gradually dry and flatten on the surface just as the skin cells on your arm do.

So technicians then follow a detailed, multi-day recipe, with dozens of different measurements that must be correct down to the microliter. Growing human cells in a petri dish is finicky work. Add an ingredient a few hours too late? Forget it, your cells are dead. But done right, those cells will replicate to form a layer 12 cells thick. Air wafts over the top cells, while the bottom layer bathes in a nutrient-rich blood substitute, much like the epidermis on your body. Ten days later, it’s Monday—the company always ships on Monday—and the skin coins are ready to ship.

Before lab-grown skin came along, the way to test whether a chemical would irritate the skin was to use bunnies. Scientist would shave off a patch of fur, smear the chemical on, and check back hours and days later. “Obviously, the more irritating the chemicals, the more gruesome it can be for the animal,” says Michael Bachelor, product manager at MatTek.

With recreated human skin like MatTek’s EpiDerm, it’s a more streamlined process. To test for irritation, you add the chemical to test along with a dye called MTT to a skin circle in a plastic well. MTT turns purple when a cell is alive. Hours or days later, a machine can measure the exact amount of dye in the well and calculate the number of living cells. The more of them are dead, the more irritating the chemical. Of course, this is not a perfect replica of what happens when you spill toilet bowl cleaner on your arm, but the MTT test is a proxy for how easily a toxic chemical can kill cells.

MatTek’s customers also use its skin tissue models to develop anti-aging creams. You don’t go looking for wrinkles in the little clear circles of cells. Instead, scientists can see how anti-aging creams turn on or off genes like those for collagen and elastin, which give skin its youthful bounce. To test anti-aging, companies use MatTek’s full-thickness skin, which includes both epidermis and the next layer of skin, dermis, because cells in the two layers of skin affect each other. These complex but poorly understood phenomena matter when you go to skin on a body.

Which brings us to the limits of current skin tissue models. Even with the full-thickness skin, it’s not exactly like skin. It doesn’t have hair follicles or nerves or oil glands. The protein scaffold on which the cells grow is simplified. MatTek’s tissue models are designed for discrete tests, where machines can look some cells and spit out a single number: This product is this much irritating or that product is that much good at killing skin pigment cells. The models work very well for these tests, but you can’t, say, graft them onto the body and expect them to start growing like skin. They do, after, look like thin clear disks of jello.