Why do dogs eat grass? What the evidence says about illness, vomiting and instinct

In a UC Davis survey of 1,571 plant-eating dogs, only about 8% regularly looked ill before grazing and 22% regularly vomited afterwards, and in a controlled Australian trial dogs ate less grass, not more, when their guts were upset.

Most dogs eat grass because grazing is normal dog behaviour, not because they are sick or short of a nutrient. In the largest study so far, an owner survey published in 2008 by Karen Sueda, Benjamin Hart and Kelly Cliff of the University of California, Davis, 68% of 1,571 plant-eating dogs ate plants daily or weekly, 92% of owners said their dog rarely seemed ill before or after, and 78% said it rarely vomited afterwards. Grass was the plant 79% of the dogs ate most.

Why dogs graze is less settled. The best-supported idea is that the habit was inherited from wolves, which also eat grass, and that it may once have helped clear intestinal worms. That last part is an untested hypothesis: dogs carrying worms in a small experiment ate no more grass, and nobody has shown that grass removes parasites from a dog.

Explain it like I’m busy:

• Grass eating is common and usually harmless: 79% of healthy dogs at a veterinary clinic had done it, and most plant-eating dogs look well before and don’t vomit after.
• No study has linked grazing to a poor diet. In controlled trials, hungry dogs ate more grass and dogs with loose stools ate less.
• The main risks are what is on the grass, such as lawn chemicals and roundworm eggs, and a grazing habit that starts suddenly or comes with other signs of illness.

Key facts about dogs eating grass

• How common: 37 of 47 healthy dogs (79%) seen for routine care at the UC Davis teaching hospital had eaten grass or other plants, their owners said.
• Illness first: 8% of 1,571 plant-eating dogs in the team’s internet survey frequently looked ill before grazing (9% in the paper’s abstract).
• Vomiting after: 22% frequently vomited within an hour; 78% rarely did.
• Diet: raw or home-fed dogs grazed no more often than dogs on commercial food (Sueda et al., 2008).
• Experiments: 12 beagles produced 374 grass-eating bouts and two vomits, and grazed less while a supplement gave them loose stools (McKenzie, Brown and Price, 2010).

What “normal behaviour” means for a grazing dog

“Normal” here means the behaviour turns up in most healthy animals, across breeds, without an obvious trigger such as pain. Grass eating passes that test: it appeared in every breed group in the UC Davis survey at similar rates, in 27 of 29 kennelled dogs watched in Australia, and in puppies five weeks old.

Normal does not mean the dog is digesting anything. Dogs can barely break down grass fibre, so blades tend to come out much as they went in. That is why a 1940s observation still shapes the debate: Adolph Murie, studying wolves in Alaska, found grass wrapped around intestinal worms in wolf droppings and suggested that grass might have a “scouring effect” that removes worms. Picture a bottle brush pulled through a tube. It is an image of a hypothesis, not a demonstrated mechanism, and the one experiment that looked for it in dogs found no effect.

How scientists know: three surveys and a kennel full of beagles

The UC Davis surveys

Sueda, Hart and Cliff tested the two claims owners repeat most: that dogs eat grass when they feel sick, and that they do it to vomit. They set a bar in advance: if three-quarters of owners said their dog rarely looked ill before grazing, the illness idea would fall, and the same rule applied to vomiting.

They asked three groups: 25 veterinary students, none of whom had seen their grazing dog look ill beforehand; owners of 47 healthy dogs at the UC Davis teaching hospital, with any dog that had gut disease, diabetes or appetite-changing drugs excluded; and, for the big numbers, an internet survey aimed at owners of plant-eating dogs.

Of 3,340 internet replies, 1,571 passed the filters: at least six hours a day with the dog, and plant eating seen at least ten times. Among those dogs, 1,439 (92%) rarely seemed ill before or after, and 1,225 (78%) rarely vomited within an hour. Both hypotheses fell below the authors’ threshold.

The Australian experiments

Surveys tell you what owners notice. To watch dogs directly, Samantha Bjone, Wendy Brown and Ian Price at the University of New England in Armidale, Australia, ran controlled experiments, reported in a 2007 conference paper, a 2010 journal paper and Bjone’s doctoral thesis.

In the first, 12 dogs were offered pots of grass three times a day for six days while their feeding time was shifted. They ate more grass before their meal than after it, and less as the day wore on, which points to hunger rather than sickness.

The second most directly tests the “upset stomach” idea. Twelve beagles (average 3.7 years, 13.3 kg) were fed a complete diet, for part of the study laced with fructooligosaccharide, a fermentable sugar that causes temporarily loose stools. If dogs graze to settle a troubled gut, they should have eaten more grass on the loose-stool days. They ate significantly less. Across 80 minutes of observation per dog there were 374 grass-eating bouts and two vomits, both from the same dog.

A third experiment, reported only in the thesis, tested Murie’s worm idea. Eighteen dogs due for deworming were offered grass, half were then dewormed, and all were tested again. Grazing time did not differ between treated dogs and those still carrying worms. Across the whole project the team logged 2,769 grass-eating bouts and 60 vomits, 47 of them in the worm study, and the dogs went back to the grass after 40 of those 60.

Show the arithmetic behind the percentages

• Internet survey: 1,439 ÷ 1,571 = 92% rarely ill, leaving 132 dogs (8.4%; the paper says 8% in its results and 9% in its abstract). 1,225 ÷ 1,571 = 78% rarely vomited, leaving 346 (22%).
• Clinic survey: 37 ÷ 47 = 79% had eaten plants; 4 ÷ 33 = 12% regularly ill first; 6 ÷ 33 = 18% regularly vomited after.
• Vomits per grass-eating bout: 2 ÷ 374 = 0.5% in the beagle experiment; 60 ÷ 2,769 = 2.2% across the whole Australian project, with 47 of the 60 vomits from the worm study.

Why do dogs eat grass and throw up?

Some do, and the UC Davis data say who. Dogs that regularly looked ill before eating plants were significantly more likely to vomit afterwards, older dogs were more likely to do both, and hound and toy breeds vomited more often. The authors’ reading is that gut discomfort sometimes prompts plant eating, and that a stomach full of grass can then come back up. Vomiting, in Hart’s phrase, looks “incidental to, rather than caused by” plant eating.

Roughly one plant-eating dog in five vomits regularly after grazing. A dog that usually brings its grass up is in the minority the UC Davis team flagged as possibly having low-level gut trouble. What the evidence does not support is the idea that dogs graze in order to vomit: the Australian dogs went straight back to the grass after being sick, which a dog using grass as an emetic would have no reason to do.

Is grass eating a sign of illness or a nutritional deficiency?

For most dogs, no. Sueda, Hart and Cliff split their 1,571 dogs by diet: 1,245 ate mainly commercial food meeting AAFCO standards, and 326 ate mainly raw or home-cooked diets, which the authors treated as more likely to be incomplete. The two groups grazed at similar rates, 67% and 71% at least weekly, a difference that was not significant. Hart added in a 2008 summary for veterinarians that dogs getting less fibre did not graze more.

The case for a deficiency rests on one dog. In 2007, Byeong-Teck Kang and colleagues at Konkuk University in Seoul described an 11-year-old Miniature Poodle that had eaten plants and vomited daily for seven years. Switched to a high-fibre diet, it stopped within three days and stayed clear for 13 months. A single uncontrolled report cannot tell you whether fibre would change anything in other dogs.

Some pet-advice sites call grass eating a form of pica, the compulsive eating of non-food items. A behaviour seen in most healthy dogs, unrelated to diet, is better described as normal, the conclusion both research groups reached. A sudden change is different. Maria Verbrugge, a clinical instructor at the University of Wisconsin School of Veterinary Medicine, put it this way in 2023: “If it’s a new behavior, it may indicate a change in your dog’s health.”

The five explanations, ranked by evidence

Inherited from wild canids (supported, mechanism unknown). Wolves eat grass. In the scat studies Sueda and colleagues summarised, plant matter turned up in about 11% of samples from Latvian wolves and about 74% of summer samples from Yellowstone, and grass specifically in 2 to 14% of scats, depending on the population. Some of that is swallowed with prey, but wolves have been seen grazing deliberately, and a 2020 review cited by Hart’s group found fibrous plant eating in 124 carnivore species. A trait shared by wild and domestic canids, every breed group and five-week-old puppies looks inherited. For what wolves eat the rest of the time, see our guide to what Arctic wolves eat.

Purging intestinal worms (proposed, untested in dogs). This is the function Hart’s group favours. Wild chimpanzees swallow rough leaves whole that sometimes emerge with worms attached, and young dogs, which carry more parasites than adults, grazed most in the UC Davis survey. The fit is suggestive, nothing more: Bjone’s deworming experiment found no change in grazing, and no one has fed dogs grass and counted the worms that came out.

Hunger (supported by one small experiment). Bjone’s 12 dogs grazed more before meals than after, so to a dog grass can simply be something to eat. The effect may be smaller in a pet fed twice a day.

Learned from the mother (partly supported). Every puppy in the 2009 study ate grass, which argues for an inborn tendency, but puppies whose mothers grazed heavily grazed more themselves.

Boredom, anxiety or attention-seeking (plausible, no data). Advice pages often list these, but SciQuest found no study that measured grass eating against boredom or anxiety in dogs.

What changes the answer

Age. Age was the only variable in the internet survey that predicted how often dogs ate plants: younger dogs grazed more and were less likely to seem ill before or vomit after. An older dog that starts grazing and vomiting is a different case from a puppy that grazes and is fine.

Who was counted, and who judged. The internet survey recruited owners of dogs already known to eat plants, so its 68% “daily or weekly” figure describes grazers, not all dogs; the 79% prevalence figure comes from 47 dogs at one hospital. Owners judged whether their dog looked ill, and the authors note that some dogs with undiagnosed gut trouble were probably included.

Which gut problem. The beagle experiment induced loose stools in the large intestine. It says nothing about nausea or stomach pain, which is where the folk theory lives, and as far as SciQuest can find no one has tested nausea directly. The controlled experiments also used only 12 to 18 dogs each.

Is eating grass safe? Lawn chemicals, parasites and toxic plants

Grass itself is not the hazard; what is on it can be. Hart’s 2008 advice to veterinarians was to keep grazing dogs away from chemically treated lawns and toxic plants, as VCA Animal Hospitals and the American Kennel Club also advise.

Lawn chemicals have been studied for a different reason. A 2012 case-control study in Environmental Research, by Biki Takashima-Uebelhoer and colleagues, compared 263 dogs with malignant lymphoma against 470 control dogs and found that owner-reported use of professionally applied pesticides was associated with 1.7 times the odds of lymphoma (95% confidence interval 1.1 to 2.7). That is an association from a questionnaire study, not proof that lawn treatment causes cancer, and it was not about grass eating.

Parasites are the better-documented risk. The Companion Animal Parasite Council notes that dogs pick up roundworms by swallowing larvated eggs from a contaminated environment, that most eggs need two to four weeks to become infective, and that they can then stay infective for years. A dog nosing through grass where other dogs have defecated can meet them whether or not it eats the grass. The council recommends year-round parasite control and regular faecal tests, a conversation for your own vet.

Finally, “grass” in a garden is rarely only grass: 21% of the surveyed dogs mostly ate other plants, such as berries and leaves, and many garden and house plants are toxic to dogs.

Why it matters to you

If your dog grazes now and then, looks well before and after, and keeps its food down, the evidence says you are looking at normal behaviour, and none of these studies gives a reason to stop it. The practical questions are about the lawn: has it been treated, do other dogs use it, and is your dog on parasite prevention?

Grazing becomes a reason to call a vet when it changes or comes with other signs: a dog that suddenly starts eating grass, one that regularly vomits after it, or one that is also off its food, losing weight, lethargic or has diarrhoea. Then the grass is a symptom to mention, not the problem to treat. For the human side of living with a dog, see why owning a pet may be good for your health.

“But wait…”

My dog always eats grass and then throws up. Doesn’t that prove the self-medication idea? It shows your dog is one of the roughly one in five that regularly vomit after grazing, and those dogs are more likely to have felt off beforehand. What it cannot show is intent. The same pattern is expected if mild nausea makes a dog swallow whatever is in front of it and a stomach full of fibre then comes back up. Discomfort and grazing go together in some dogs; nobody has shown which causes which.

Related questions

Why do puppies eat grass? Because almost all dogs do, and young ones most. All the five- to seven-week-old puppies in Bjone’s study ate grass, and in the UC Davis survey the youngest dogs grazed most often and were least likely to be ill or to vomit.

Do dogs eat grass when they are hungry? In the one controlled test, yes: dogs ate more grass before their meal than after it. Walking your dog after breakfast is a low-cost way to see whether that applies to yours.

Do cats eat grass for the same reason? In a 2021 survey of 1,021 cats by Hart’s group, 71% had been seen eating plants at least six times, 9% seemed ill beforehand and 27% frequently vomited afterwards. The authors proposed the same inherited explanation, untested in cats too. Our explainer on why cats purr covers another feline behaviour that is less settled than it looks.

In one sentence

Dogs eat grass because grazing is a normal, probably inherited habit: in the largest survey only about 8% of 1,571 plant-eating dogs looked ill first and 22% regularly vomited after, and the idea that it once helped clear worms is untested.

Quick check: did it click?

1. What share of the 1,571 plant-eating dogs in the UC Davis survey regularly looked ill before grazing? About 8% (rounded to 9% in the paper’s abstract); 92% rarely did.
2. What happened to grass eating when beagles were given loose stools with a supplement? They ate significantly less grass, not more.
3. Has anyone shown that eating grass clears worms from a dog? No. It is a hypothesis based on wolf and chimpanzee observations, and a deworming experiment found no change in grazing.

Read next

• Why do cats purr? How the sound is made and what it means
• Why big dogs die younger: DNA study finds faster epigenetic ageing
• Why owning a pet may be good for your health

Disclosure: This article was drafted with AI assistance and checked against the sources below. The percentages and ratios in the “Show the arithmetic” box are SciQuest calculations from the published counts, and the survey chart is an original SciQuest graphic drawn from Sueda, Hart and Cliff (2008). We read the full 2008 paper, Hart’s 2008 veterinary summary and Bjone’s doctoral thesis, which contains the 2010 beagle experiment and the unpublished worm study; the 2007 conference paper was checked through the University of New England’s record of its abstract. This article is not veterinary advice. Spotted an error? See our corrections policy.

Sources

1. K. L. C. Sueda, B. L. Hart and K. D. Cliff (2008), Characterisation of plant eating in dogs, Applied Animal Behaviour Science 111, 120 to 132 (pilot, clinic and internet surveys; 79%, 68%, 8 to 9%, 22%; age, breed, diet and vomiting analyses; wolf scat summary; parasite hypothesis)
2. B. L. Hart (2008), Why do dogs and cats eat grass?, Veterinary Medicine, December 2008, 648 to 649 (fibre finding; “incidental to, rather than caused by”; advice on treated lawns and toxic plants)
3. S. J. Bjone, W. Y. Brown and I. R. Price (2007), Grass eating patterns in the domestic dog, Canis familiaris, Recent Advances in Animal Nutrition in Australia 16, 45 to 49 (12 dogs; more grazing before meals and earlier in the day)
4. S. J. Bjone, W. Y. Brown and I. R. Price (2009), Maternal influence on grass-eating behavior in puppies, Journal of Veterinary Behavior 4, 97 to 98 (abstract; puppies 5 to 7 weeks old)
5. S. McKenzie, W. Y. Brown and I. R. Price (2010), Reduction in grass eating behaviours in the domestic dog, Canis familiaris, in response to a mild gastrointestinal disturbance, Applied Animal Behaviour Science 123, 51 to 55 (12 beagles; fructooligosaccharide; 374 bouts, 2 vomits; read in the thesis version below)
6. S. Bjone (McKenzie) (2009), Grass-eating behaviour in the domestic dog, Canis familiaris, doctoral thesis, University of New England, Armidale (observational study of 29 dogs; worm-burden experiment; project totals of 2,769 bouts and 60 vomits)
7. B.-T. Kang, D.-I. Jung, J.-H. Yoo, C. Park, E.-J. Woo and H.-M. Park (2007), A high fiber diet responsive case in a Poodle dog with long-term plant eating behavior, Journal of Veterinary Medical Science 69, 779 to 782 (single case)
8. B. L. Hart, L. A. Hart, A. P. Thigpen and N. H. Willits (2021), Characteristics of plant eating in domestic cats, Animals 11, 1853 (open access; cat survey figures; Murie 1944; Franck and Farid 2020 review of 124 species)
9. D. R. Stahler, D. W. Smith and D. S. Guernsey (2006), Foraging and feeding ecology of the gray wolf (Canis lupus): lessons from Yellowstone National Park, Journal of Nutrition 136, 1923S to 1926S (vegetation in summer scats; 74% figure as summarised by Sueda et al.)
10. B. B. Takashima-Uebelhoer et al. (2012), Household chemical exposures and the risk of canine malignant lymphoma, Environmental Research 112, 171 to 176 (263 cases, 470 controls; professionally applied pesticides OR 1.7, 95% CI 1.1 to 2.7)
11. Companion Animal Parasite Council, Ascarid guidelines (infection from larvated eggs in the environment; 2 to 4 weeks to become infective; survival for years; year-round control)
12. University of Wisconsin School of Veterinary Medicine (2023), Dog’s weirdest snack: grass (Maria Verbrugge quotation)
13. VCA Animal Hospitals, Why do dogs eat grass? and American Kennel Club, Why does my dog eat grass? (veterinary guidance on treated lawns, parasites and warning signs)