Labrador Retriever: What the UK Records Actually Show

This page is not veterinary advice. Paw Multiverse is not a veterinary practice. We cannot examine, diagnose or treat your animal. There are no doses here, no portion sizes, no feeding amounts, no exercise durations and no treatment steps, on purpose. If your animal has changed in a way you cannot explain, ring your vet or your nearest emergency clinic.

This page is built on two documents and nothing else. The first is the Royal Kennel Club breed standard for the Labrador Retriever, last updated 1 April 2025, read on 5 August 2026. The second is Labrador retrievers under primary veterinary care in the UK: demography, mortality and disorders, a VetCompass study published on 22 October 2018 and read on the same day.

One describes the dog the breed clubs are aiming at. The other counts what walked through the door of a British veterinary practice. They do not agree, and the place where they disagree is the most useful thing on this page.

What the standard asks for

The standard is short. It wants a dog that is “strongly built, short coupled, very active”, with a coat that is “short dense without wave or feathering, giving fairly hard feel to the touch” over a “weather-resistant undercoat”. On character it asks for a dog that is “intelligent, keen and biddable, with a strong will to please”, with a “kindly nature, with no trace of aggression or undue shyness”.

On size it gives an ideal height at the withers of 56 to 57 cm for dogs and 55 to 56 cm for bitches. On colour it is unusually specific: the only correct colours are wholly black, yellow and liver or chocolate, yellows running from light cream to fox red, and “any other colour is unacceptable”. Three colours, listed as equals.

That last sentence is the one to hold on to.

What 33,320 Labradors actually showed

The VetCompass study took every Labrador Retriever under primary veterinary care in the UK in 2013, 33,320 dogs, and read the clinical records of a random sample of 2,074 of them. This is not a survey of owners and it is not a show ring. It is what a vet wrote down.

  Labrador Retriever, as the VetCompass study records it
  =====================================================
  study population    33,320 dogs, UK, 2013
  clinical sample     2,074 dogs
  median longevity    12.0 years (IQR 9.9 to 13.8)
  commonest disorder  otitis externa 10.4% (95% CI 9.1 to 11.8)
  second              overweight or obesity 8.8% (7.6 to 10.1)
  third               degenerative joint disease 5.5% (4.6 to 6.6)
  commonest death     musculoskeletal 24.5%, neoplasia 21.1%
  adult bodyweight    males 35.2 kg (SD 5.9), females 30.4 kg (SD 5.2)
  published           22 October 2018

Read the top of that list again. The two most commonly recorded problems in the breed are an ear condition and being too heavy. Neither is dramatic, neither is genetic destiny, and both are the sort of thing that is noticed by somebody who is paying attention to the animal week by week.

The colour finding

The standard treats black, yellow and chocolate as three equally correct colours. The clinical records do not.

Median longevity for chocolate Labradors in the study was 10.7 years (IQR 9.0 to 12.4). For everything else it was 12.1 years (IQR 10.2 to 13.9), and the difference reached statistical significance at P equal to 0.028. Otitis externa was recorded in 23.4 per cent of chocolate dogs, 17.0 per cent of yellow dogs and 12.8 per cent of black dogs, with P less than 0.001. Pyo-traumatic dermatitis, the inflamed patch of skin often called a hot spot, was recorded in 4.0 per cent of chocolate dogs against 1.1 per cent of black dogs, P equal to 0.011.

What that finding is: an association measured in one national dataset in one year. What it is not: a mechanism. The authors of the study did not establish why, and neither will we. The most commonly offered explanation, that a smaller breeding pool for a recessive colour concentrates other traits, is plausible and is not something we can source to a publisher, so it is not asserted here.

The honest use of this figure is narrow. It does not mean a chocolate Labrador in front of you is ill. It means that if you are choosing, you now know something the breed standard does not tell you.

How long they live, and what ends it

Median longevity across the whole breed was 12.0 years, interquartile range 9.9 to 13.8, range 0.0 to 16.0. Among the deaths where a cause was recorded, musculoskeletal disorder accounted for 24.5 per cent and neoplasia for 21.1 per cent. Those two together are close to half of everything.

A musculoskeletal cause of death in a large dog usually means the animal reached a point where it could not get up. That is the same axis as the third most common disorder in the list, degenerative joint disease at 5.5 per cent. It is worth noticing that the commonest disorder and the commonest cause of death in this breed are connected by weight, which is the second commonest disorder. We are not going to tell you what to feed or how much: see the food reference for how this site handles that question.

Bodyweight, and what we refuse to do with it

The study gives median adult bodyweight of 35.2 kg for males, standard deviation 5.9 kg, and 30.4 kg for females, standard deviation 5.2 kg. The breed standard gives an ideal height and no weight at all.

We publish those numbers because they are what a national dataset recorded. We will not turn them into a target for your dog, and we will not publish a body condition score chart, because deciding whether a particular animal is the right weight is a clinical judgement made with hands on the animal. A number read off a web page is the wrong tool for that job.

Figures we will not reproduce

Two things this site declines to print, and the reason in each case.

Any feeding amount. Sources exist that publish daily quantities by bodyweight. We do not reproduce them, because a quantity that is right for one dog is wrong for the next one and the reader has no way to tell which they have.
Any exercise duration. The same. A figure in minutes reads like a prescription and this site is not in a position to write one.
Any price. There are no prices anywhere on this site. The cost pages, including the Labrador cost page, explain what the categories of spending are and where published figures can be found, without inventing a number.

What we could not source

Stated plainly, because a gap named is more useful than a gap filled with guesswork.

Why chocolate dogs did worse. No publisher we reached offers an established mechanism. The study reports the association and stops there, and so do we.
A UK life table figure for this breed. The 2022 UK dog life tables give a life expectancy at age zero for the Labrador Retriever of 11.77 years, but the confidence interval was not available in the material we could read, so we report the VetCompass median instead and flag this.
Hip and elbow scheme distributions. We could not obtain a current, dated, publicly readable distribution of hip and elbow scores for the breed, so no scoring figure is quoted here.
Anything about training. Not a sourcing failure. Training is a procedure and this site does not publish procedures.

Where to read next

The other breed pages on this site are built the same way, from a breed standard and whatever national clinical data exists, with the disagreements left visible: Golden Retriever, German Shepherd Dog, French Bulldog, Beagle, Dachshund, Rottweiler and Yorkshire Terrier.

If you have not settled on a breed, the breed selector asks about your circumstances rather than your preferences. For the money side, see the cost hub and the vet costs guide, and for what the same question looks like across countries, the country comparison.