Horses in a field with an animal caretaker in Val de Loire INRAE research centre, France. (Credit: INRAE - Christophe Maitre)
In A Nutshell
- Horses detect human fear through scent alone. French researchers exposed horses to cotton pads soaked with human sweat from people watching horror films versus comedy clips.
- Fear odors changed horse behavior and physiology. Horses smelling fear-scented pads startled more intensely, avoided human contact 40% more, and showed elevated heart rates during stress tests.
- Joy odors had minimal effect. Horses exposed to joy-related human sweat behaved similarly to horses exposed to neutral control odors, suggesting fear signals cross species boundaries more readily than positive emotions.
- Handler anxiety may spread to horses through sweat. The findings suggest that nervous riders or trainers could inadvertently increase horse reactivity through chemical signals their bodies produce during stress.
Turns out the old saying about animals smelling fear isn’t just folk wisdom. Horses actually can detect when humans are afraid through scent alone, and new research suggests that fear-related emotional states may be shared across species from person to animal through chemical signals in sweat.
Scientists in France exposed 43 horses to cotton pads soaked with human armpit sweat collected during moments of fear or joy. Horses smelling fear-scented pads became measurably more fearful themselves. They startled more violently when surprised, stared longer at unfamiliar objects, kept their distance from humans, and showed elevated heart rates during stressful moments. The study, published in PLOS One, offers evidence that human emotional odors can influence horses’ behavior and physiology in ways consistent with emotional responses.
For anyone who handles horses, this matters. A nervous rider or trainer doesn’t just communicate anxiety through body language or voice. Their sweat may carry chemical markers that horses detect and respond to, potentially creating a feedback loop where human nervousness amplifies horse nervousness.
Capturing Fear and Joy in Cotton Pads
The research team at INRAE in Nouzilly, France, first had to capture authentic emotional odors. They recruited 30 volunteers who watched either a 20-minute horror film excerpt or joyful comedy clips while wearing cotton pads under their armpits to absorb sweat.
After each viewing, participants rated their emotional experiences. The researchers selected samples from the 14 people who experienced the strongest fear during horror clips and the strongest joy during comedy clips. Those samples were frozen at negative 70 degrees Celsius until test day, alongside unused control pads.
How the Horses Responded
Researchers fitted horses with lycra muzzles containing two cotton pads positioned directly in front of their nostrils. Each horse underwent four behavioral tests: a grooming session, free interaction with a standing human, a surprise umbrella popping open while eating, and exposure to an unfamiliar object.
The differences were clear. Horses exposed to fear odors touched the experimenter about 40% less frequently during free interaction compared to horses in the joy group. Even during grooming, a typically pleasant activity, fear-scented horses showed reduced social contact.
Startle Responses and Physiological Changes
The most dramatic results appeared during the suddenness test. Researchers placed food pellets on the ground and waited for each horse to start eating. Three seconds later, an assistant triggered a folded umbrella to suddenly pop open above the bucket.
Horses in the fear odor group startled nearly twice as intensely as horses in the control group. Their maximum heart rates during the startle were also significantly higher than horses smelling joy or control odors, showing a physiological response consistent with heightened arousal or fear. However, the researchers note that heart rate data had measurement issues requiring substantial data cleaning, and salivary cortisol levels showed no significant differences between groups.
When exposed to an unfamiliar object made from colorful plastic pieces, horses smelling fear odors gazed at it significantly longer, a classic sign of wariness in horses.
Interestingly, horses exposed to joy odors didn’t behave dramatically differently from horses exposed to neutral control odors. Fear signals appear to cross species boundaries more readily than positive emotions, possibly because fear could serve as an early warning system about danger.
Questions About How and Why
The researchers note that their observations are consistent with emotional contagion, where an emotional state transfers between individuals, but emphasize that independent replication is needed to confirm this interpretation. The specific chemical compounds responsible for these effects remain unknown, as the study did not analyze the chemical composition of the odor samples.
Why might horses have this ability? Several possibilities exist. If animals can detect fear-related odors across species, they could gain advance warning about threats they haven’t directly observed. Scent molecules linger after the source has left, travel around obstacles, and function in darkness, making them effective warning signals.
Whether horses developed this sensitivity through thousands of years of domestication or inherited it from wild ancestors is unclear. Dogs show similar responses to human stress odors, raising questions about whether this ability is widespread among mammals or specific to species with long histories of human interaction. The researchers suggest that mammals might share receptors inherited from a common ancestor, but this remains speculative.
What This Means for Riders and Trainers
These findings carry weight for anyone in the equestrian world. Riders preparing for competitions, therapeutic riding instructors working with nervous clients, and veterinarians handling anxious horses might inadvertently influence horse behavior through odors associated with their emotional states.
The research team notes that if handler anxiety affects horse reactivity through olfactory channels, managing handler stress becomes a professional competency affecting everyone’s safety. Recognizing this pathway of communication could change how equestrian programs approach training.
The old horseback riding advice to “stay calm so the horse stays calm” may have a chemical component. Horses aren’t just reading posture, voice, or facial expression. They appear to be detecting something in human odors and responding to it, though exactly what they’re detecting and how it affects them requires further research.
Paper Notes
Study Limitations
The investigation included only Welsh mares aged 5-12 years, which means results may not generalize to males, geldings, other breeds, or horses of different ages. The testing protocol used an audience horse present during all tests to prevent social isolation stress, though researchers took precautions to minimize her influence by using the same audience horse across groups and habituating her to all stimuli.
Heart rate data required substantial cleaning due to measurement artifacts, resulting in exclusion of 10-28% of recordings depending on the test. Individual data points differing by more than 35 beats per minute from the previous point were removed as artifacts, and if more than 30% of a horse’s data for a given test was artefactual, all data for that test were excluded.
Salivary cortisol measurements showed high variability and no significant differences between groups. Post-hoc power analysis revealed a power of only 0.15, indicating the sample size was insufficient to detect effects given the variability, possibly due to circadian rhythm effects or the relatively short test duration.
The human odor donors were predominantly female (12 of 14), and individual differences in odor production related to sex, genetics, age, or physical condition were not directly analyzed, though donor identity was included as a random factor in statistical models to account for such variations. The study did not analyze the chemical composition of the odor samples to identify specific compounds responsible for the observed effects.
Funding and Disclosures
This study received funding from ANR Emodour (grant number ANR-23-CE20-0033) and Institut Français du Cheval et de l’Équitation (IFCE, grant number 32001331 Cognition Emotion). Plotine Jardat received support from the French Ministry for Research. The funding sources had no role in study design, data collection and analysis, decision to publish, or preparation and submission of the manuscript.
The authors declared no competing interests. The study received ethical approval from the Val de Loire Ethical Committee (CEEA VdL, Nouzilly, France, authorization number CE19-2022-1511-2) and complied with French and European guidelines for animal research. The horses were not food deprived and did not undergo invasive procedures.
Human participation was carried out according to the Declaration of Helsinki and was approved by the Institutional Review Board of the University of Tours (authorization number 2022-029). All participants provided written informed consent. The authors disclosed using ChatGPT and Mistral AI tools to improve readability and language, reviewing and editing all content afterward.
Publication Details
Title: Human emotional odours influence horses’ behaviour and physiology | Authors: Plotine Jardat, Alexandra Destrez, Fabrice Damon, Noa Tanguy-Guillo, Anne-Lyse Lainé, Céline Parias, Fabrice Reigner, Vitor H. B. Ferreira, Ludovic Calandreau, Léa Lansade | Affiliations: Institut Français du Cheval et de l’Equitation, Pôle développement, Innovation et Recherche, Nouzilly, France; INRAE, CNRS, Université de Tours, PRC, Nouzilly, France; Development of Olfactory Communication and Cognition Laboratory, Centre des Sciences du Goût et de l’Alimentation, Institut Agro Dijon, CNRS, Université de Bourgogne-Franche-Comté, Inrae, Dijon, France; UEPAO, INRAE, Nouzilly, France | Journal: PLOS One | Publication Date: January 14, 2026 | Volume/Issue: 21(1): e0337948 | DOI: https://doi.org/10.1371/journal.pone.0337948 | Received: July 24, 2025 | Accepted: November 14, 2025 | Data Availability: Data and code are publicly available at https://doi.org/10.57745/T2RTDR







