Can Kennel Cough Go Away on Its Own? (And Can It Kill a Dog?)

Last reviewed · 21 September 2026 · Sources linked throughout

Usually, yes. Most dogs with kennel cough get better without antibiotics and without anything dramatic happening. The international guidelines that vets themselves work from say so plainly — most dogs stay bright, keep eating, and come right on their own. What matters is knowing the small number of signs that mean this dog is the exception, because those signs are specific, they arrive together rather than one at a time, and they are the difference between a cough you can wait out and an illness that needs a vet today.

What “on its own” actually means — and what it doesn’t

The clearest statement on this comes from the International Society for Companion Animal Infectious Diseases antimicrobial guidelines (Lappin et al., J Vet Intern Med 2017;31:279–294), which are written for vets deciding whether to reach for antibiotics at all:

“The majority of cases of CIRDC are currently believed to be viral in etiology and so antimicrobial administration is often not indicated. Most dogs with clinical signs of CIRDC including mucopurulent nasal discharge maintain normal appetite and attitude and might resolve spontaneously within 10 days without antimicrobial treatment.”

Read that carefully, because the referent matters. It says most dogs recover without antibiotics. It does not say kennel cough needs no care, and it does not say you should never involve your vet. Rest, water, a harness instead of a collar, and keeping your dog away from other dogs are still the job — they just aren’t a prescription. “It’ll sort itself out” and “it doesn’t need antibiotics” are different sentences, and only the second one is supported.

Note the ten days as well. That is the guidelines’ observation window for reconsidering antibiotics, not the length of the illness. A kennel cough usually runs one to three weeks from start to finish, and the cough is often the last thing to go. Our day-by-day recovery timeline covers what a normal course actually looks like.

The four signs that say it isn’t resolving on its own

Here is the part worth committing to memory. The same guidelines say antibiotics should be considered inside that ten-day window “only if fever, lethargy, or inappetence is present together with mucopurulent discharges”. The MSD Veterinary Manual describes the escalation the same way — as a cluster, not a checklist:

“Development of more severe signs, including fever, purulent nasal discharge, depression, anorexia, and a productive cough, is indicative of bronchopneumonia.” (MSD Veterinary Manual, Kennel Cough)

Both sources use “together with” and “including” rather than “or”, and that is not an accident. A dog with a loud cough and a bit of clear nasal discharge who is still mugging you for dinner is behaving exactly as an uncomplicated case behaves. The picture that changes the decision is a dog who is off — flat, not eating, feverish — and producing thick discoloured discharge. One item from the list is worth a phone call. Several at once is worth an appointment.

Two of those you can check at home, using the same numbers we use everywhere on this site:

• Temperature. 38.3–39.2 °C (101.0–102.5 °F) is normal. 39.2–39.5 °C is above normal but not yet fever. 39.5 °C (103.1 °F) and above is a fever — contact your vet.
• Resting breathing rate. Count breaths for a full minute while your dog sleeps. 15–30 is normal. Sustained above 30, or more than 20% above your dog’s own usual number, means contact your vet. Above 35–40, or visible effort at any rate, means same-hour. How to count it properly takes two minutes to learn.

The breathing rate is the more useful of the two, because it tracks the thing that actually matters — whether the infection has moved from the windpipe down into the lungs.

Can kennel cough kill a dog?

It can, and it very rarely does. That sentence is unsatisfying, so here is the evidence behind both halves of it, with the denominators stated, because this is a subject where numbers get quoted without them.

The European review of CIRDC describes Bordetella infection as high-morbidity and “rarely fatal”, and is explicit about the route by which it does turn serious: “In puppies and dogs with immunosuppression or other concurrent diseases, CIRDC can be complicated by bronchopneumonia, resulting in more severe signs such as dyspnoea, weight loss, pyrexia and even death” (Day et al., J Comp Pathol 2020;176:86–108). Kennel cough does not kill dogs by being a bad cough. It kills, when it kills, by becoming pneumonia. (That review’s advisory board is sponsored by Zoetis, a vaccine manufacturer — worth knowing, though nothing in the passage above is a claim about a product.)

On how often: the same review draws on a UK health survey of purebred dogs which recorded 15,881 deaths, of which respiratory disease of every kind accounted for about 1.2% — pneumonia specifically 0.3%, with laryngeal paralysis, choking, bronchitis and tracheal collapse making up most of the rest (Adams et al., J Small Anim Pract 2010;51:512–524). A second survey of 5,663 dogs recorded no respiratory deaths at all. The review’s own conclusion: “respiratory disease is rarely lethal in the pet dog population.”

One caution about that 1.2%, since it is the kind of figure that gets repeated wrongly. It is a share of deaths, not a risk per dog — roughly one death in eighty in that survey was attributed to a respiratory cause. It covers every respiratory disease, not kennel cough. And it comes from owner-reported questionnaires returned by breed-club members, with a 24% response rate. It tells you respiratory illness is an uncommon way for a dog to die. It is not a kennel cough survival rate.

The most directly relevant number is from the worst-case group. A retrospective series looked at 65 puppies under a year old, all hospitalised, all with confirmed community-acquired infectious pneumonia — dogs who had already crossed the line this whole page is about. 57 of the 65 (88%) survived to discharge. Bordetella bronchiseptica was the organism in 32 of them (49%), and those puppies had it harder than the rest: younger (median 14 weeks against 21), more often needing supplemental oxygen (25 of 32 against 16 of 33), and longer in hospital (mean 7.2 days against 4.9) (Radhakrishnan et al., J Am Vet Med Assoc 2007;230(10):1493–1497). Eight of those puppies did not go home, and that is the honest other half of the figure.

What does not exist is a case-fatality rate for kennel cough itself. A search of PubMed, Google Scholar, JAVMA, JSAP and the Journal of Veterinary Internal Medicine across 1990–2026 turns up no population-level figure for ordinary household dogs, and the reason is structural: most dogs with kennel cough are never counted anywhere, because they never see a vet, so there is no denominator to divide by. Anyone quoting you a percentage is quoting a study of hospitalised dogs and calling it something else. What the evidence supports is the shape rather than the number: ordinary kennel cough in a healthy adult dog is not a mortal illness, and even among the puppies sick enough to be admitted with pneumonia, most went home.

A second, sicker cohort is worth putting beside it. A Glasgow teaching-hospital series reviewed 49 dogs of all ages admitted with pneumonia of any cause — not just infectious, and not just kennel cough. 33 of the 49 (67%) survived to discharge (Conway et al., J Vet Emerg Crit Care 2021;31(4):490–497). We give you both numbers rather than the friendlier one, because they measure different populations: puppies with community-acquired infection do better than a referral hospital’s full pneumonia caseload, which includes aspiration and dogs with other illnesses underneath. Neither is a kennel cough statistic. Both describe dogs already far sicker than the one most people are reading this about.

If pneumonia is the thing you are actually worried about, our guide to telling pneumonia from kennel cough is the one to read next.

Which dogs are the exception

The literature keeps naming the same groups, and they are worth knowing because the advice on this page changes for them. Don’t wait and watch if your dog is:

• A puppy. Less respiratory reserve, more exposure, and the group that dominates the pneumonia case series. If your dog is under about six months, the threshold for calling is lower — see kennel cough in puppies.
• Elderly, or living with another illness. Heart disease, chronic airway disease, anything affecting the immune system.
• On immune-suppressing medication — steroids, chemotherapy, some skin and joint drugs.
• Flat-faced. Bulldogs, pugs, French bulldogs and their relatives start with less airway to spare, so the same infection costs them more.
• Recently in a group setting — a shelter, a busy kennel, a rescue transport. Not because those dogs are frailer, but because mixed infections and heavier exposure are more likely.

For everyone else — a healthy adult dog, vaccinated or not, bright between coughing fits — watchful waiting is a reasonable plan, and it is roughly what your vet would suggest on the phone.

It’s been three weeks and it hasn’t gone

This is the second question people arrive with, and it usually has one of three answers.

The cough is outlasting the illness. This is common and not alarming on its own. Airway lining damaged by infection stays irritable after the infection has cleared, so a dog can be entirely well and still cough at excitement, at a pull on the lead, or first thing in the morning. Cornell’s veterinary centre notes the same pattern in canine influenza specifically: “Dogs may continue to cough for several weeks following recovery from acute infection.” A cough that is fading, in a dog who is otherwise themselves, usually just needs more time.

It was never only kennel cough. A cough that has not improved at all after two to three weeks deserves a fresh look, because several conditions get mistaken for a kennel cough that won’t shift — tracheal collapse in small breeds, a heart-related cough in older dogs, chronic bronchitis, lungworm, or something inhaled. These are diagnoses, not guesses, and they need an examination rather than another week of waiting.

It got better and then got worse. A cough that improves and then relapses, or a dog who perks up and then goes flat again, is the pattern worth taking seriously. Go back to the four signs above.

What helps while you wait — and what doesn’t shorten it

Being straight about this is more useful than a list of remedies. No controlled study has shown that anything available to you at home shortens kennel cough. A search of the veterinary literature from 1990 to 2026 turns up no trial that measured illness duration for any of the things owners are routinely told to do — not harnesses, not rest, not steam. What supportive care does is make your dog more comfortable while their own immune system does the work, and stop you making things worse. That is worth doing. It is just not the same as treatment.

• Swap the collar for a harness while the cough lasts. This one has a clear mechanism behind it even without a trial: the MSD manual notes that in kennel cough “the cough is easily induced by gentle palpation of the larynx or trachea” — so a lead pulling on an inflamed windpipe sets off a fit. Fewer fits, more comfort.
• Rest, and no vigorous exercise. Short lead walks rather than the park. Same logic: less coughing, not faster healing.
• Keep the air moist. Sitting with your dog in a steamy bathroom can ease a dry, hacking cough. It doesn’t treat anything; it is comfort, and it is free.
• Keep them away from other dogs. Not for your dog’s benefit — for everyone else’s. How long kennel cough stays contagious covers how long that needs to last, and it is longer than most people expect.
• Don’t reach for human cough medicine. Several ingredients that are routine for people are dangerous for dogs, and cough suppression is actively the wrong move when a cough is wet and productive — that cough is clearing the airway. What your dog does and doesn’t need goes through this properly.

And antibiotics? Not usually, for the reason at the top of this page — most kennel cough is viral. What is striking is how thin the evidence is even for the cases where antibiotics are given. The most recent randomised trial in dogs with infectious tracheobronchitis compared four antibiotics against each other in 60 dogs and opens by stating that “therapy with antibiotics in canine ITB still remains questionable” (Haydardedeoğlu et al., J Adv VetBio Sci Tech 2023;8(3):210–215). Every dog in it received an antibiotic; there was no untreated arm. So it can tell a vet which antibiotic works fastest once they have decided to prescribe one — and it cannot tell anyone whether prescribing at all shortens an uncomplicated case. That question has not been answered in a controlled trial. It is a decision your vet makes against the cluster of signs above, and when your dog really needs antibiotics sets out how.

Frequently asked questions

How long should I wait before ringing the vet?

You don’t need to wait at all if you are worried — a phone call costs nothing. As a rule of thumb: a bright dog with a dramatic cough can be watched. A dog with a fever, who is off food, flat, or breathing faster than usual at rest should be seen. And a cough that has not started to improve after ten days is worth a conversation regardless.

Will my dog get better faster with antibiotics?

Not if the cause is viral, which it usually is. The guidelines are explicit that antibiotics are often not indicated in kennel cough, and are for the minority of cases with signs of bacterial complication.

Can kennel cough turn into pneumonia?

Yes — that is the complication that makes it serious, and it is much more likely in puppies, elderly dogs, and dogs whose immune systems are compromised. The warning signs are fever, thick discoloured discharge, lethargy, loss of appetite and a wet, productive cough, arriving together.

My dog is vaccinated. Can they still get it?

Yes. The vaccine covers Bordetella and sometimes parainfluenza, but kennel cough is caused by a whole group of organisms. Vaccinated dogs generally get a milder, shorter illness — see kennel cough in vaccinated dogs.

Does a dog need to be isolated until the cough stops completely?

Not necessarily — a lingering post-infection cough is not the same as still being infectious, and the two do not end on the same day. Our contagion guide covers the timings.

Where to read next

• Signs of kennel cough — how to recognise it and judge severity
• What kennel cough looks like — the visual guide
• What kennel cough sounds like — a sound-by-sound guide
• What CIRDC actually is — the full picture behind the nickname