The word "pheromone" appears on roughly half the fragrance ads you will encounter online. It also appears in a significant body of peer-reviewed neuroscience research. The two usages have almost nothing in common. Here is what the actual science has established, what it has not, and why the distinction matters.

What a pheromone actually is

In biology, a pheromone is a chemical compound produced by an organism that triggers a specific, predictable behavioral or physiological response in another member of the same species. The key word is specific. A pheromone does not just smell interesting - it reliably causes a defined response. The queen substance in honeybees, which prevents worker bees from developing ovaries, is a textbook example.

This kind of rigid chemical determinism works well in insects, where nervous systems are simpler and behavioral repertoires are more constrained. In mammals, chemical signaling exists but is considerably less deterministic. A mouse pheromone triggers territorial behavior or mating readiness - but context and prior experience modify the response.

In humans, the question becomes even more complicated. For a true pheromone system to function, you need the sensory equipment to detect it. That equipment is the vomeronasal organ.

The human VNO problem

The vomeronasal organ (VNO) is a small structure in the nasal cavity that, in many animals, detects chemical signals and sends them directly to brain regions controlling reproductive and social behavior. In humans, the VNO is present - you can find it anatomically - but the neural wiring that makes it functional in other animals is largely absent.

Human embryos develop a functional VNO early in gestation, but it appears to lose its nerve connections before birth. By adulthood, it is what anatomists call a vestigial structure: present but without apparent function.

This does not mean humans cannot detect and respond to chemical signals from other people. It does mean the classical pheromone pathway - VNO to hypothalamus to behavior - is not the mechanism. Whatever chemical communication humans engage in appears to run through the regular olfactory system, which is a different circuit with different properties.

What research has actually found

Several compounds have been studied as candidate human pheromones. The most researched are androstadienone (AND), found in male sweat and seminal fluid, and estratetraenol (EST), found in female urine.

A 2007 study by Wyart and colleagues, published in the Journal of Neuroscience, found that exposure to androstadienone significantly increased cortisol levels in women compared to a control condition - a real physiological effect. The same researchers noted effects on focus and mood. This was a carefully controlled study and the finding has been replicated in some subsequent work.

A notable 2005 study by Savic and colleagues in the Proceedings of the National Academy of Sciences used brain imaging to show that AND and EST activated different hypothalamic regions in heterosexual men and women, with some specificity by sexual orientation. The hypothalamus is involved in reproductive regulation, which made this significant.

But "significant" in science means real and worth studying, not large and reliable. Effect sizes in pheromone research tend to be small. Replication across labs has been inconsistent. Context - whether participants knew what they were smelling, whether they found the scent pleasant - substantially modified results. None of this resembles the deterministic insect pheromone system.

The McClintock effect

In 1971, researcher Martha McClintock published a study in Nature suggesting that women living together in dormitories showed menstrual cycle synchrony - their periods tended to converge over time. The paper became one of the most cited in the field and was widely interpreted as evidence of human chemical communication.

Subsequent attempts to replicate the finding produced mixed results. A rigorous statistical analysis by Yang and Schank in 2006 argued that the apparent synchrony could be entirely explained by mathematical chance - given how periods work, some degree of overlap is statistically expected regardless of any social signaling.

The debate continues, but the strongest methodological verdict is that menstrual synchrony as evidence of human pheromone communication is not well supported. This does not mean chemical communication between people does not happen - it means this particular example is shaky evidence for it.

What actually drives attraction

Scent does play a genuine role in human attraction - but through a different mechanism than pheromones. The strongest evidence comes from research on major histocompatibility complex (MHC) genes, which regulate immune response.

A series of studies beginning with Wedekind and colleagues in 1995 found that people tend to prefer the body odour of individuals with different MHC profiles from their own. This preference was detected through actual olfaction - smelling worn T-shirts. The proposed mechanism is that genetic diversity in immune genes produces offspring with broader immune coverage. The finding has been replicated and is considered reasonably robust.

This is chemical signaling, and it is consequential. But it operates through the regular olfactory system, involves genetic information rather than behavioral commands, and is probabilistic rather than deterministic. It is one input among several - alongside appearance, voice, social context, and the accumulated texture of familiarity.

Physical attraction, in other words, is a weighted sum of cues. Scent is genuinely one of them. A pheromone bottled and sold is not.