Most breed pages have one hard chapter. This one has two, and both are unusually well measured. That is uncomfortable to read. It is also why the page is worth your time, because each comes with a screening scheme that has a name, a written protocol, and a certificate you can ask to see.
If you already share your house with one of these dogs, read on anyway. Half of what follows is about what to watch for and when to ask for a heart check.
Size, and where the panel’s numbers come from
Say this plainly. The figures above are not measured from a large group of pet Cavaliers. They are worked back from the breed standard, and our dataset flags them as medium confidence for that reason.
The American standard asks for “Height 12 to 13 inches at the withers; weight proportionate to height, between 13 and 18 pounds.” It sets no separate figure for males and females. The split in the panel is our estimate, using the median gap between the sexes across the breeds in our data where real weights were recorded. Treat it as a sensible expectation rather than a measurement.
Being small does bring one advantage. The largest growth-standard study in dogs excludes only dogs whose adult weight passes 40 kg, about 88 lb. A Cavalier sits nowhere near that line, so our chart works inside the range it was checked on. It also means the large-breed cautions you have read elsewhere are not aimed at this dog. Controlled calcium, slow growth, keeping a puppy off stairs: those come from research in retrievers and mastiffs.
Temperament, and what the breed name actually buys you
Start with the finding that governs every breed description ever written, including this one. When researchers put owner reports next to DNA in a 2022 study of thousands of dogs, the breed a dog belonged to accounted for roughly a tenth of how it behaved. The other nine tenths sat elsewhere: in the individual animal, its upbringing and its life. A pedigree is a weak forecast.
That said, this is a breed people choose for character rather than looks, and the character claim survives contact with evidence.
The best Cavalier-specific behaviour data we could open comes from a 2026 survey of 9,402 dogs, which included 715 Cavaliers. It scored them on the C-BARQ, a validated owner questionnaire with twelve behaviour scales. The comparison was against Cavapoos, and the Cavaliers came out lower (meaning better) on eight of the twelve. Those eight included all three aggression scales: toward the owner, toward strangers and toward other dogs. They also covered fear of strangers, fear of other dogs, general fearfulness, separation problems and excitability.
The Cavapoos beat them on one thing. Trainability.
Read that carefully. It is a comparison with one crossbreed, not a ranking against every dog, and the sample recruited itself online. It is still the largest set of behaviour scores published on this breed.
What we will not claim
You will find pages saying this breed scores among the least aggressive in the classic C-BARQ breed comparison. We went and looked. That paper names Golden Retrievers, Labradors, Bernese Mountain Dogs, Brittanys, Greyhounds and Whippets at the low end. The Cavalier is not in that list, so we will not report a placing it does not give. The same study checked its own recruitment, which is worth knowing. Owners found through breed clubs and owners found online gave only small differences in aggression scores, and the two samples ranked the breeds in much the same order.
The breed standard asks for a dog that is “Gay, friendly, non-aggressive with no tendency towards nervousness or shyness.” That is an ideal a breed club wrote down, not a measurement of living dogs.
When behaviour is actually pain
One warning belongs here rather than in the health section. A Cavalier that starts yelping for no reason, scratching at its neck, or ducking away from a collar may be in pain rather than being difficult. Among dogs affected by syringomyelia, about 45% first showed signs in their first year and a further 40% between one and four.
The puppy year
Adolescence hits this breed on the usual timetable, and it is documented well enough that you can stop blaming yourself.
Guide-dog puppies were tested on a sit command at five months and again at eight. At eight months they obeyed less, and only for the person raising them. Handed to someone they knew less well, the same dogs did better. The rudeness is aimed at you specifically. A separate British project followed puppies from six to eighteen months, and complaints from owners topped out at twelve months, where 42.1% had at least one thing to report.
Neither study had Cavaliers in it, so take the timing rather than the detail. What is breed-specific is what else the first two years are for. A Cavalier can have its first MRI scan graded from twelve months, and its first heart certificate from eighteen.
The heart, which is the centre of this breed
Myxomatous mitral valve disease (MVD, where the valve between the two left chambers of the heart thickens and stops sealing) is why this page exists. Blood leaks backwards, and a stethoscope hears that leak as a murmur.
In UK primary-care records covering 1,749 Cavaliers, a heart murmur was recorded in 541 of them, 30.9%. It was the most common single thing recorded in the breed, ahead of diarrhoea, dental disease and ear infections. The sample was a young one, with a median age of 4.0 years at first visit and 5.25 at last. What the study does not tell you is how old those 541 dogs were when the murmur was heard.
| Where it was measured | Dogs | What was found |
|---|---|---|
| UK vet records, all ages | 1,749 | murmur in 30.9% |
| UK breed heart scheme | 8,860 (16,887 exams) | murmur first heard at a median of 7.8 years in males, 8.3 in females |
| UK breed-club screening, dogs aged 8+ | 126 | murmur in 89%; valve disease on scan in 100% |
That last row deserves unpacking. In 126 Cavaliers aged eight or over, every dog had mitral valve disease on an ultrasound scan. Those dogs were recruited through regional breed clubs and seen at shows and in owners’ homes, not pulled off a clinic waiting list. But 44.4% of them had no enlargement of the heart at all. Having the disease and being ill from it are not the same thing, and the gap between them is where a long life happens.
UFAW’s review of the evidence describes the share of Cavaliers with a murmur rising by roughly 10 percentage points a year, so about half have one by five and nearly all by ten. The same review records that in most breeds a leaking mitral valve turns up from six or seven years onward. In Cavalier males it can be heard from three to four, and in females four to five. The disease is not unique to the breed. Arriving three years early is.
What a murmur grade does and does not tell you
If your vet has just said “grade two”, here is what that means. Grade is loudness, on a scale of one to six. It is not a stage of illness.
The ACVIM consensus guidelines sort dogs into stages instead. Stage B1 is a dog with a murmur whose heart has not changed shape. Stage B2 is a dog whose left atrium and ventricle have enlarged past set measurements. Murmur intensity of 3 out of 6 or louder is one of four things used to reach B2: the other three need a heart scan and a chest X-ray. The panel names this breed as predisposed to MVD young, and says owners of breeding dogs in high-risk breeds may choose to join yearly screening events. So a grade alone is a prompt for imaging, not a verdict.
The reason the stage matters
Reaching B2 changes what treatment does. The EPIC trial randomised 360 dogs with a murmur of grade 3 or louder and an enlarged heart. Of the 354 dogs it went on to analyse, 161, or 45.5%, were Cavaliers, the largest single breed in the trial. Half got pimobendan, a drug that helps the heart squeeze harder and relaxes the vessels it pumps into. Half got a placebo.
Median time until heart failure, cardiac death or being put to sleep for heart reasons: 1,228 days on the drug against 766 days on placebo. The hazard ratio was 0.64, with a confidence interval of 0.47 to 0.87: the risk at any moment was roughly a third lower. That is about fifteen extra months, and it is the strongest argument for imaging a murmur rather than watching it.
Heart disease is also what shortens this breed. In the UK Kennel Club and BSAVA purebred survey, cardiac causes were behind 42.8% of Cavalier deaths.
The breeding schemes, and what the evidence shows
This is the part almost nobody reports, and it is the most useful thing on the page.
There is an agreed protocol, usually called the MVD breeding protocol. It asks that no Cavalier be bred before 2.5 years of age, that it have a clear heart then, and that both its parents were still murmur-free at five years. The UK breed club’s version asks for parents’ clear certificates “issued at 5 years of age or older”. Sweden’s version, compulsory since 2001, lets a dog breed from 24 months if it and its parents are clear, and otherwise makes it wait until four. All of it is set out in UFAW’s review.
In the UK the paperwork comes from the Kennel Club Heart Scheme. A dog can be tested from 18 months, with repeats at 2, 4 and 6 years. The certificate carries a murmur grade and a prolapse grade for how far the valve bulges on the scan. Those combine into a traffic light. Green is “lowest risk of developing clinical MVD and the lowest risk of passing the condition on to any offspring.” Red dogs “should not be used for breeding.”
Does any of it work? Three studies, one long enough to answer
This gets written up as a league table of countries. It is not one. Nobody has compared these schemes head to head. Three separate teams measured different dogs, at different ages, over different stretches of years, and the stretch of years is what decides whether a result means anything.
Start with what is not the difference between them. Denmark’s scheme and Sweden’s both began in 2001, and both are mandatory for breeders. So neither the start date nor compulsion separates the two findings below.
Denmark: a real change, measured across a decade. Researchers looked at 1,380 cardiac examinations of 997 dogs between 2002 and 2011. They compared 215 dogs examined in 2002–03 with 172 examined in 2010–11 whose parents had gone through the scheme. Every group had a median age of about two to two and a half years, so like was compared with like. The odds of a murmur in the scheme’s offspring were 0.27 against the baseline, a confidence interval of 0.16 to 0.43. That is a 73% cut. Dogs of the same age whose parents had skipped the scheme showed no improvement.
Hold on to one detail. Those dogs were around two years old. What dropped was how often a murmur turned up that early, which is precisely the trait the scheme selects against. It is not a 73% cut in the disease over a whole life, and the paper does not say it is.
Sweden: nothing found, in a window too narrow to find anything. Look at the design rather than the result. 131 dogs were examined at six years old: one group born in 2001, the other born in 2003. Murmur prevalence sat around 52 to 55% and did not shift. But the scheme only started in 2001, so the parents of both groups were mostly mated before it existed. Two birth years apart is not a decade apart. The authors make the point themselves, writing that the programme “might have a slow effect on disease prevalence over a longer period”. That is a null result from a short window, not a scheme that failed.
The UK: a small shift, and a compliance problem. The breed scheme’s own records cover 16,887 examinations of 8,860 dogs across twenty years. Bitches seen by general-practice vets picked up their first murmur later over time, moving from 8.6 to 9.2 years. No change appeared in males. The number that explains the rest: in the final five-year block, only 4% of dogs had been bred in line with the advice.
So the honest conclusion is a narrow one. The Danish study is the only one whose window was long enough to show a change, and it showed one. Sweden’s window was two birth years wide and could not have shown much either way. Twenty years of UK records mainly show that hardly anyone followed the advice. None of that ranks one country against another. It does say that a scheme actually followed for a decade moved something real.
What to ask for
Ask to see both parents’ heart certificates, and look at the age on them, not just the word “clear”. A clear certificate at 18 months means little in a breed where murmurs arrive at four. One issued at five or older is the one worth having. Ask how old the parents were when mated, and whether the grandparents were screened.
Syringomyelia: two numbers that get confused constantly
Chiari-like malformation, usually shortened to CM, is a skull that is small for the brain inside it, so the back of the brain gets crowded at the opening where the spinal cord leaves. Syringomyelia, or SM, is what can follow: fluid-filled cavities inside the spinal cord itself.
Here are the two figures. They measure completely different things.
On MRI. Researchers scanned 555 Cavaliers whose owners said they had no symptoms. Syringomyelia was present in 25% of dogs at 12 months, rising to 70% at 72 months and over. Only age predicted it; sex made no difference. A Dutch and Belgian screening programme found the same pattern in 1,249 scans of 1,020 dogs. The malformation itself was near-universal: 1,225 of 1,243 graded scans, 98.6%, were CM2, the most affected grade. A syrinx big enough to score SM2 was present in 195 of 1,243, about 16%, rising from 11% under three years to 32% over five.
In the clinic. Reviewing owner surveys, UFAW found the share of Cavaliers actually showing signs ranged from none up to about 35%, with one US questionnaire landing at 3.4%. British veterinary records tell a similar story: clinical syringomyelia affects about one Cavalier in 60, against one dog in 2,000 across all breeds. Of those affected, 72% were recorded as being in pain.
Both numbers are real. “Seventy percent of Cavaliers have syringomyelia” and “about one in sixty is treated for it” do not contradict each other. The first counts what a scanner sees; the second counts dogs that got ill. Anyone quoting one without the other is telling you half of it.
The scheme, and what to ask for
The BVA and Kennel Club run a CM/SM scheme that grades an MRI scan. Two grades come back on one certificate, plus a letter for the dog’s age band: a for over five years, b for three to five, c for one to three.
| Grade | Chiari-like malformation | Syringomyelia |
|---|---|---|
| 0 | No Chiari malformation | Normal, no dilation and no syrinx |
| 1 | Cerebellum indented, not rounded | Central canal dilation under 2 mm |
| 2 | Cerebellum impacted into or herniated through the opening at the rear of the skull | Syrinx of 2 mm or more |
The Kennel Club publishes a mating table setting out which graded dogs may sensibly be paired, and advises against breeding from any dog showing signs. Screening is suggested at about a year, at three to five, and again past five.
So the request to a breeder is short: CM/SM certificates for both parents, with grades and age letters. An SM0 at one year is a far weaker statement than an SM0 at five, which is exactly what the age letter exists to tell you.
Two conditions you can simply test away
This is the cheerful section.
Two inherited disorders in this breed were mapped in a single 2012 study. Episodic falling syndrome, sudden bouts of stiffness and collapse brought on by exercise or excitement, comes from a large deletion in a gene called BCAN. Dry eye curly coat syndrome, which affects the tear glands and the skin from birth, comes from a single missing base in FAM83H.
Across 122 dogs sampled, the faulty version of each gene turned up at a frequency of 0.08, roughly one gene copy in twelve. Both are recessive. A puppy needs two faulty copies to be affected, a carrier with one is perfectly healthy, and two carriers bred together is the only way an affected puppy arrives. Because both have a definitive DNA test, a breeder can rule both out. “Clear” or “carrier” on both, for both parents, is a complete answer.
Other things worth raising at a check-up
Loose kneecaps. In 210,824 dogs in English practices, patellar luxation (a kneecap slipping out of its groove) was diagnosed in 1.30% overall, at a median age of 4.0 years. Cavaliers were the fourth most common breed among the cases, contributing 52 of them, 6.1%. Read that cautiously: a share of cases partly reflects how common a breed is, not only its risk. The signal to watch is the skipping step a small dog takes for a stride or two and then shakes off.
Glue ear. Primary secretory otitis media is a plug of thick mucus behind an intact eardrum. It matters here because it causes head and neck pain, head tilt, scratching and partial deafness: signs that overlap heavily with syringomyelia. Material has been found in the middle ear of 28 to 54% of scanned Cavaliers, though nobody has measured it across the whole population. Most affected dogs respond to having the ear flushed. If your dog has neck pain, name this one out loud. It is the treatable explanation.
How long they live
In UK primary-care records covering 102,609 dogs and 5,095 confirmed deaths, Cavaliers had a median longevity of 9.9 years from 124 deaths, with the middle half falling between 8.1 and 12.3. Every dog in that same study, of every breed, came in at 12.0 years. So this breed sits about two years short, and that is a comparison made inside one dataset rather than stitched between two. For what it is worth, a separate UK study of 584,734 dogs puts dogs generally at 12.5 years, which points the same way.
Why is not mysterious. Cardiac disease accounted for 42.8% of deaths in the breed’s own health survey. Across all breeds in the primary-care study, the leading causes were tumours at 16.5%, musculoskeletal problems at 11.3% and neurological ones at 11.2%: a spread no single condition dominates. In Cavaliers, one does.
Claims we could not stand behind
A month-by-month Cavalier puppy weight chart. The tables in circulation cite nothing. Our own figures come from a model fitted to breed adult weights. That model sits inside its validated range at this size, but it is a prediction, not a record of what real Cavalier puppies weighed.
Any daily exercise figure. Not in the breed standard, not in any study we could open.
That a particular breeder’s line is free of either condition. Nobody can show you this. The malformation behind syringomyelia was near-universal in the largest screened group, and valve disease reached every dog in a group of aged Cavaliers. Certificates describe individual dogs at an individual age. They do not clear a bloodline.
This breed’s own row in the 2024 life-table paper. The per-breed estimates live in a supplementary file we could not open, so we have used the older primary-care figure and said which is which.
Breed-specific hearing loss figures. The nearest study measured hearing electrically in 20 Cavaliers graded for Chiari-like malformation. Dogs with the worse grade had slower signals along the hearing nerve, but the authors called the effect “not always statistically significant or predictable.” Twenty dogs is not a prevalence.
Anything about children. No study grades any breed for this. It is not a measurable quantity, here or on any other page we write.