In 1995, the Swiss zoologist Claus Wedekind asked a group of women to sniff the unwashed t-shirts of several anonymous men and rate how pleasant they smelled. To ensure the women were getting the real deal, Wedekind had the men sleep in the shirts for two nights prior, avoiding deodorants, alcohol, cigarettes, smell-altering foods, scented soaps, and anything else that could cover up or neutralize their natural scents. As unappealing as it sounds, this now-famous experiment was not a study of cruelty. In many ways, it was a study of love.

Before the experiment, scientists had already discovered that many animals use their sense of smell to find compatible mates, naturally avoiding partners who can’t provide the healthiest offspring. But Wedekind’s experiment offered provocative insight into the question of whether humans do this as well. He found that each woman had strong preferences among the dirty shirts, overwhelmingly favoring the smell of men who were more genetically compatible with her. In other words, the t-shirts that smelled the most pleasant—sometimes even reminding women of former partners—came from men with whom the women could, genetically speaking, produce healthy babies.

But if humans do find mates through scent, this study presents a paradox. How many people ever wear shirts they’ve slept in for two days on first dates? Natural odor is often the last thing anyone wants to share with a potential mate. Indeed, entire commercial industries exist to mask it: Deodorizers, soaps, shampoos, perfumes, breath mints, and the like are often marketed explicitly as tools of attraction. Their whole appeal is that they’re supposed to cover unsexy bodily stenches and turn heads in the office elevator.

Are these products attracting the wrong people? Bathing in intoxicating odors may offer one type of allure, but what if doing so sacrifices a more fundamental mechanism of human seduction? If body odors contain basic genetic information, then obscuring them conceivably could risk throwing off people looking for a good match.

Wedekind’s notion of genetic compatibility is based on what scientists refer to as the Human Leukocyte Antigen, or HLA. This group of genes essentially acts as security for the immune system. The HLA make proteins that attach themselves to suspicious cells in the body––a pathogen, perhaps, but even cancer cells––and present the cells to the immune system to determine whether they are safe.

What do these genes have to do with attraction? “The most important thing for determining health is your immune system,” explains Rachel Herz, the author of The Scent of Desire. From a biological perspective, she says, the perfect mate for producing robust progeny would have a fairly dissimilar HLA to one’s own—“fairly” being an intentional distinction here. Too dissimilar an HLA and certain immunities may not be inherited. Too similar, and a potentially hazardous mismatch occurs, one that Herz explains has been linked to “spontaneous abortions, miscarriages, and issues with recessive birth defects.”

Though not universal, by and large people naturally gravitate towards compatible mates—specifically, mates who aren’t related to them. One study of a closed religious sect called the Hutterites demonstrated that even within a limited gene pool, members still generally select partners with dissimilar immune genetics.

It’s unclear if smell played any role in helping the Hutterites find compatible mates so successfully. It is clear, though, that body odor is undesired in modern culture. If smell remains involved in detecting compatible mates, there would seemingly need to be some aggressive biological reinforcements to overcome daily showers, deodorizers, chewing gum, and perfume.

One possible reinforcement is pheromones—odorless chemicals released to attract or influence mates. Since they can’t be smelled, these chemicals would just need to be inhaled to have an effect; colognes and deodorants wouldn’t stop them. Pheromones have been confirmed in other species, like mice and, perhaps most famously, honeybees. But there’s no hard evidence humans even produce them (though that doesn’t stop some believers from attending pheromone parties). There’s evidence that people possess the olfactory hardware needed to detect pheromones, yet scientists disagree on how and if this organ is actually functional.

Another possibility is that perfumes and other manufactured scents don’t end up competing with a person’s natural odors in the first place. There are a handful of studies that show women tend to select scents that don’t mask their HLA, but resemble it. This is rather profound. What this research suggests is that people may prefer fragrances that smell like them, thus amplifying, not neutralizing, their genetics. It’s a nice idea, but some researchers, like Rachel Herz, are “not very confident about the data.”

Perhaps the most obvious answer to the smell paradox, however, is that compatibility is determined by more than odors. Even if fragrances do mislead potential mates, there may be enough other clear signals of compatibility to compensate. Daniel Geraghty, the president of Scisco Genetics, is less convinced by the t-shirt studies––which he says could have many interpretations—than by the population genetic evidence like the Hutterites, which he says demonstrates that people are often able to “recognize, somehow, whether it’s through smell or some other different means, the difference in the immune response genetics.”

Geraghty’s laboratory specializes in sequencing immune genes like the HLA for disease research and organ transplantation. To sequence HLA genetics, they use a variety of cells––saliva, blood, and so on. The research is a reminder that couples are likely to interact with each other’s DNA in many different ways. Kissing, for one, might present a significant opportunity for a couple’s HLA to get together and decide if the relationship should progress. Not every culture kisses romantically, but nearly every culture—even the Hutterites—has a version of kissing: touching lips, rubbing noses, smelling cheeks. This swapping of scents and microbes may exist as a follow-up to the smell test.

Even if perfumes and deodorizers do confuse some initial attraction, in other words, there simply seems to be no hiding one’s DNA when a couple begins to share a bed. A true disruption of humans’ ability to read or relay genetics would likely have to be more fundamental than scent-altering products—something that effectively alters the natural signals and receptors used to detect mates. This may in fact exist: Some drugs do affect how humans produce and perceive odor, which could potentially have an impact on mate detection when the drug is in use. Though more research would need to be done, Wedekind’s studies, for instance, have shown that women on birth control often select less genetically compatible mates than those who were not on an oral contraceptive, supporting the idea that, at least with smell, certain drugs may indeed confuse matters.

That said, the whole idea that humans choose mates based on immune genetics is still firmly in the realm of speculation, propped up by a series of compelling studies. HLA is a particularly complex part of the genome to research, and geneticists like Daniel Geraghty are really just beginning to crack into its potential. “The list is long for what HLA could be used for,” says Geraghty. “What it is used for is much less than that.”

In organ transplantation, HLA matching ensures a host does not reject a transplant (siblings are generally preferred donors because they are most likely to share HLA). For that use, compatibility is significant. But the links between HLA mismatches and birthing complications have not been strong enough for HLA matching to be widely considered when diagnosing issues of infertility or recurrent pregnancy loss. As the Kentucky Fertility Institute in Louisville told me, “until more data on the clinical utility of HLA typing becomes available, it is not something we would routinely factor into our management considerations.”

There are likely many couples who are, in fact, genetically incompatible. Genetics, after all, are far from the only factor that determines how or why people end up in relationships. Kindness comes to mind. Style, sense of humor, shared interests, and intelligence may also be weighed heavily when picking someone with whom to buy a home, share a bank account, and raise children.

If nothing else, the evidence does suggest that humans have the capacity to choose compatible mates based on how pleasant they smell. So if you’re specifically looking for someone to make a baby with, it still may not be a bad idea to follow your nose.