Horses really can smell our fear, new study finds

Here is a deep dive into the methodology, statistical findings, and strict controls used in the January 2026 study published in PLOS One. 

The Experimental Subjects

• The Donors: Sweat samples were collected from 30 human volunteers. To prevent chemical contamination, they followed strict protocols: no alcohol, no spicy foods, and no deodorants.
• The Equine Sample: The researchers tested 43 Welsh mares (average age of 7.9 years). They lived in outdoor paddocks at the INRAE center in Nouzilly, France.
• Social Support: To prevent separation anxiety from altering the data, an "audience horse" was kept nearby as a calm companion during all tests. 

Eliminating Scent Contamination

To prove horses react only to human emotional chemosignals, the researchers eliminated all external odors: 

• The lead researcher, Plotine Jardat, took two showers a day using unscented soap.
• Assistants wore FFP2 masks, swimming caps, and clean overalls changed for every single horse.
• The testing box was thoroughly blasted with a water jet between subjects.
• The sweat-infused cotton pads were secured directly inside disposable Lycra muzzles fitted to the horses' nostrils.

---

The Three Behavioral and Physiological Tests

The horses were randomly divided into three groups: Fear scent, Joy scent, and Control (unused cotton pads). They underwent three specific trials: 

1. The Human Interaction Test (Grooming & Approach)

• Setup: A familiar handler groomed the horse and then stood slightly apart.
• Data: Horses wearing the "fear sweat" muzzle pulled away. They touched the human significantly less, showing a Rate Ratio (RR) of 0.60 compared to the other groups.

2. The Suddenness Test (Reactivity)

• Setup: While a horse was eating from a food bucket, an experimenter suddenly popped open an umbrella.
• Data: The "fear" group experienced an intense startle reflex (Effect size: Cohen's d = -0.88). Their maximum heart rate spiked dramatically compared to horses smelling joy or neutral pads (Effect size: Cohen's d = 1.16). 

3. The Novelty Test (Investigation)

• Setup: Horses were presented with a strange, abstract sculpture made of linoleum, plastic, and string.
• Data: Instead of approaching and sniffing the object, the "fear" group kept their distance and simply stared at it with heightened vigilance (Rate Ratio of 1.32 for gazing). 

---

The Cortisol Paradox

Interestingly, while the heart rates and behaviors perfectly mirrored a state of panic, the horses' salivary cortisol levels (the primary stress hormone) did not show a statistically significant difference between the groups. This suggests that "smelling fear" triggers an immediate, acute sympathetic nervous system reaction (fight-or-flight) rather than a slow-building, long-term endocrine stress response. 

Would you like to explore how dogs compare to horses in smelling human stress, or should we look at the practical training techniques riders can use to mask or manage their anxiety around horses?