Can dogs smell stress? What a breath-and-sweat experiment found

The study came from Clara Wilson and colleagues at Queen’s University Belfast, published in PLOS One in September 2022. They recruited volunteers, took a sample of each person’s breath and sweat, then gave them a nasty mental task: count backward from 9,000 in steps of 17, out loud, for three minutes, while a researcher sternly corrected mistakes.

Right after the task, they took a second sample. Each was a gauze pad wiped on the back of the neck plus three deep breaths blown into a vial. Crucially, they only kept a person’s samples if the stress was real: self-reported stress had to rise by at least two points on a 10-point scale, and where it was measured, heart rate and blood pressure had to go up too. Of 53 people sampled, 36 made the cut, and the dogs sniffed their samples within three hours.

The sniffers were four ordinary pet dogs from the Belfast area, including a cocker spaniel, a cockapoo and two mixes, trained with a clicker and kibble. Twenty dogs started training; most dropped out or didn’t meet the bar.

In each test the dog faced three sniffing ports: the person’s post-task sample, the same person’s calm sample, and a blank. The handler didn’t know which port held the stress sample. The dogs alerted to the stress sample in 675 of 720 trials, or 93.75 percent, against a chance rate of one in three. Individual dogs scored between 90 and 96.88 percent.

The detail that impressed the researchers most: on the very first trial with a new person’s samples, before any repetition could teach them that individual’s smell, the dogs chose correctly 34 times out of 36. The authors concluded that an acute stress response changes the odor of our breath and sweat, and that dogs can detect the change.

Nobody knows yet. The study didn’t identify the chemicals. The authors’ working idea is that stress shifts the mix of volatile organic compounds our bodies give off, through fast changes in breathing and sweating. You will see articles say dogs smell cortisol, the stress hormone. The paper itself is more careful: cortisol peaks 10 to 30 minutes after a stressor, while these samples were taken immediately, so something faster may be involved.

It is also worth being clear about what this study did not show. It tested acute stress from a lab task, not sadness, grief or long-running anxiety. The dogs were trained to find a target smell and rewarded for it, which is different from a pet noticing your mood on its own. And four dogs is a proof of principle, not a survey of all dogs. The authors say all of this in the paper.

A 2024 study suggests it might. Zoe Parr-Cortes, Nicola Rooney and colleagues at the University of Bristol, publishing in Scientific Reports, ran 18 dogs through a cognitive bias test, a standard way to read an animal’s mood. Dogs learned that a bowl in one spot held a treat and a bowl in another spot was empty, then were shown bowls in ambiguous spots in between.

When the room carried the breath and sweat of an unfamiliar stressed person, dogs were slower to approach the ambiguous bowl nearest the empty spot, which the researchers read as a more pessimistic state. The smell of a relaxed person had no such effect. “Working dog handlers often describe stress travelling down the lead,” Rooney said, and she argues the study shows stress can travel through the air as well.

Two details make this more interesting. The odor donors were strangers to the dogs, which hints at a stress signal shared across people rather than a dog knowing one person’s smell. And, oddly, dogs learned the original treat-or-empty locations faster when the stress smell was present. Rooney has said the findings matter for dogs living in kennels and for how companion and working dogs are trained.

A separate 2019 Linköping University study found that long-term stress levels, measured by cortisol in hair, tended to rise and fall together in 58 dogs and their owners, with the owner’s personality mattering more than the dog’s. That study shows a link, not that smell causes it.

None of these studies tested a daycare lobby or an exam room, so treat this as sensible inference rather than proven fact. But if dogs can detect the odor of acute stress, and that odor can tilt their mood, then your nerves are part of the environment at the moments that already worry many dogs.

• Keep drop-offs short and boring. A long, anxious goodbye at the dog daycare door gives your dog more time to read you. Hand over the leash, say a cheerful word and go. Staff see this every morning.
• Do your worrying before you leave. If a first boarding stay is making you anxious, pack the night before and settle the details with the kennel in advance. Our guide to boarding an anxious dog covers the practical steps.
• At the vet, breathe first. Take a minute in the car before you go in. A slower, calmer you won’t erase the smell of the clinic, but it is one less signal piling on.
• Don’t blame yourself. Stress happens, and dogs live with it fine. The point is to stack the odds, not to fake serenity.

Noses are clearly doing more work in a dog’s life than we notice. For two more examples from the bathroom end of things, see why dogs kick after pooping and the debated north-facing poop study.