The Canine Ledger

Does Diarrhoea Need Antibiotics? What the Evidence Says

Antibiotics are often given to dogs with sudden diarrhoea. The trials and a 2024 European review say most dogs don't benefit. When they do.

Written by thecanineledger

This page carries no advertising, and nothing on it is for sale.

If your dog has diarrhoea and is also vomiting repeatedly, weak, collapsed, has a swollen or painful belly, or is passing a lot of blood, phone your vet or an emergency clinic now. This page is about the dog who is otherwise bright. → Vomiting and Diarrhoea: When It Can Wait and When It Can’t

In 2020, a team in Munich gave sixteen dogs with sudden, uncomplicated diarrhoea either a common antibiotic, amoxicillin-clavulanic acid, or a placebo. The dogs on the antibiotic got better no faster than the dogs on the dummy tablet. What changed was in their gut. During treatment, the share of E. coli in their stools that was resistant to the antibiotic rose to a median of 100%, and three weeks later it was “still increased” (Werner et al., J Vet Intern Med 2020).

The authors concluded that the antibiotic “confers no clinical benefit to dogs with AD, but predisposes the development of amoxicillin-resistant E. coli, which persist for as long as 3 weeks after treatment.”

That is one small trial. It is not alone. While there are dogs with diarrhoea who need an antibiotic, and quickly, the evidence now says that for the ordinary bout of sudden diarrhoea in an otherwise well dog, an antibiotic usually doesn’t help. This page sets out what that evidence is, where it is strong, where it is weak, and what to ask your vet.

1. Why this question matters

Antibiotics are a familiar answer to diarrhoea. The Merck Veterinary Manual’s guide for owners says they “are commonly given daily by mouth for several days until recovery is apparent, although their effectiveness in treating digestive system disease is still uncertain. In fact, overdoses or prolonged use of antibiotics can be harmful to the digestive tract” (Merck, Introduction to Digestive Disorders of Dogs).

In practice they are still used a lot. A 2025 study of UK veterinary records looked at 1,835 dogs seen for sudden diarrhoea in 2019. Antibiotics were recorded in 38.2% of them (O’Neill et al., PLOS ONE 2025). The authors concluded that “antibiotic use remained frequent, despite years of recommendation to the contrary.” The same study found that more than 80% of the dogs needed only one visit, which says something about how most cases go.

There are two costs to an antibiotic that doesn’t help. The first is to your dog: side effects, and disruption to the gut bacteria. The second is shared. Every unnecessary course gives resistant bacteria a chance to spread, and the resistant E. coli in the Munich trial were in the dogs’ stools, which end up on lawns, pavements and shoes.

2. What the best evidence says

In 2024, a European network of veterinary specialists set up to improve antibiotic use, ENOVAT, published a systematic review of every randomised trial it could find on treating sudden diarrhoea in dogs (Scahill et al., Vet J 2024). It found six trials of antibiotics and six of probiotics and similar products. It sorted dogs into mild, moderate and severe disease, “based on the presence of systemic signs and response to fluid therapy.”

Its central finding:

“High certainty evidence showed that antimicrobial treatment did not have a clinically relevant effect on any outcome in dogs with mild or moderate disease.”

“High certainty” is the top grade on the scale the review used. It means the reviewers think further research is very unlikely to change the answer. For the sickest dogs, the picture is different: “Certainty of evidence was low for dogs with severe disease.” That isn’t evidence that antibiotics don’t work in severe cases. It means there isn’t enough good research to say.

The same network then turned the review into a treatment guideline (Jessen et al., Vet J 2024). Its summary says the panel drafted “Four strong recommendations informed by high certainty evidence, and three conditional recommendations informed by very low or low certainty evidence.” We could read only that summary. The full guideline is behind a paywall and a bot check, so we haven’t quoted what each recommendation says.

Our note: We read only the abstracts of the ENOVAT guideline and its systematic review, so we don’t state the recommendations themselves.

3. The trials behind it

Four of the trials are worth knowing about, because they show the shape of the evidence.

Bloody diarrhoea without sepsis (Unterer et al., 2011). Sixty dogs in hospital with sudden bloody diarrhoea, and no signs of blood infection, got amoxicillin-clavulanic acid or a placebo. The researchers found no difference in deaths, time in hospital or severity of illness. Their conclusion was cautious: “In some dogs with HGE that show no signs of sepsis, antibiotics may not change the case outcome or time to recovery” (Unterer et al., J Vet Intern Med 2011).

Uncomplicated diarrhoea (Werner et al., 2020). The Munich trial above: sixteen dogs, no clinical benefit, and resistant E. coli for weeks.

Metronidazole versus placebo (Langlois et al., 2020). This is the trial that cuts the other way, and it deserves a fair hearing. Thirty-one dogs with sudden diarrhoea of no identified cause got metronidazole, another common antibiotic, or a placebo. The diarrhoea cleared sooner on metronidazole: an average of 2.1 days against 3.6 days (Langlois et al., J Vet Intern Med 2020). The authors still didn’t recommend it routinely. They wrote that more studies were needed “because most dogs achieve resolution of diarrhea within several days regardless of treatment.”

Metronidazole versus a probiotic (Stübing et al., 2024). Twenty-seven dogs got metronidazole or a probiotic product. There was no meaningful difference in how fast they recovered, but the antibiotic changed their gut bacteria. The authors concluded that “metronidazole negatively impacts the microbiome without affecting clinical outcomes” (Stübing et al., Vet Sci 2024).

These are small trials, of 16 to 60 dogs. The systematic review is what gives them weight together. Even read as generously as possible, the best case for an antibiotic in a mild case is that it might shorten the diarrhoea by about a day and a half, at the cost of resistant bacteria and a disrupted gut.

Our note: “About a day and a half” is our summary of the Langlois result (mean 2.1 vs 3.6 days).

4. When an antibiotic is the right call

None of this means antibiotics are never needed. The Merck Veterinary Manual’s chapter on sudden bloody diarrhoea, revised in 2025 and citing the ENOVAT guideline, draws the line clearly: “Parenteral antimicrobials are not recommended in mild to moderate cases” (Merck, Acute Hemorrhagic Diarrhea Syndrome in Dogs). (“Parenteral” means by injection or drip.) It then says antibiotics are suggested where there is concern for severe disease, for example when a dog still shows signs of whole-body illness after fluids, or when blood tests suggest infection is spreading.

The Merck owner guide makes the same distinction. Antibiotics “may be given by injection when septicemia (blood poisoning) is apparent or likely to occur.”

In plain terms, the dogs who may need antibiotics are the very sick ones, and the decision is made with an examination and often blood tests. That decision belongs to your vet. A puppy with suspected parvovirus is a separate case again, and an emergency. → Parvovirus: Survival, Hospital Care, and the Outpatient Option

Our note: Section 4 describes “severe” in owner terms, from Merck’s professional criteria.

5. What about probiotics?

Probiotics are often offered as the gentler alternative. The ENOVAT review looked at those too, and found that probiotics and similar products “did not show a clinically significant effect in shortening the duration of diarrhoea (based on very low to moderate certainty evidence).”

So the honest answer is that the evidence for probiotics in sudden diarrhoea is weak, and the best of it shows little effect. One small trial found a probiotic did no worse than metronidazole, but that’s a comparison with a drug that itself did little. We don’t recommend any product.

Our note: The two sources point slightly different ways on nutraceuticals.

6. What to ask your vet

If your dog has mild diarrhoea and is otherwise well, and your vet doesn’t prescribe an antibiotic, that is in line with the evidence. It isn’t the vet doing less.

If your vet does prescribe one, it is reasonable to ask, politely:

  • What makes you think my dog needs it? There are good answers, such as signs of a more serious infection.
  • What should I watch for that would change things?
  • What should I do if it isn’t better in a few days?

Don’t ask for an antibiotic because a previous dog was given one. And don’t give your dog leftover antibiotics from an earlier illness, or from a person. Never give any medicine your vet hasn’t prescribed for this dog, for this illness.

Most sudden diarrhoea in otherwise well dogs settles within days. Merck’s owner guide notes that “diarrhea can actually be a defense mechanism for the animal, helping it to eliminate harmful organisms and their toxins.” What matters most is spotting the dog who isn’t settling, or is getting worse. → Vomiting and Diarrhoea: When It Can Wait and When It Can’t

Our note: These questions are ours. Asking them doesn’t imply your vet is doing something wrong.

7. What we could not verify

Commonly saidWhat we found
“Diarrhoea needs antibiotics to clear it”Not for most dogs. A 2024 systematic review found high-certainty evidence of no clinically relevant effect in mild or moderate cases.
“Antibiotics can’t hurt, just in case”They can. In one trial, resistant E. coli rose to a median of 100% during treatment and stayed raised three weeks later.
“Metronidazole is the standard treatment”Common, but not well supported. One trial found it shortened diarrhoea by an average of about a day and a half; another found no clinical benefit over a probiotic.
“Probiotics fix diarrhoea”Weak evidence. The same review found no clinically significant effect, on very low to moderate certainty evidence.
What the ENOVAT guideline recommends, word for wordNot read. The full text was behind a paywall and a bot check. We read its abstract and the review it was built on.

Sources

Always confirm with your vet before acting on anything on this page.