Why the Companion Dog Registry Sets a 10 % COI Target for Puppies
What the coefficient of inbreeding actually measures—and what it doesn't.
By: Carolyn Kelly
When the breed group lines up at a dog show—identical shiny coats, identical gaits—it’s easy to see why people love predictability in dogs. Perfect heads, matching top-lines, stunningly beautiful dogs. But behind the scenes hides a numbers tug-of-war that breeders argue about all the time: how much inbreeding is too much, and what’s the best way to measure it?
One camp leans on the coefficient of inbreeding (COI), which tells you how much of a pup’s DNA came from the same ancestor on both sides of the family tree. Another camp pushes diversity scores—heterozygosity tests that count how many markers are different rather than duplicated. The stakes aren’t academic: when either number drifts the wrong way, whole breeds can slide toward heart disease, cancer, or fertility trouble years before problems show up in the whelping box.
Understanding those two yardsticks—and the debate around them—helps us decide whether we’re genuinely safeguarding canine genetics or letting problems build beneath the surface.
The Companion Dog Registry requires predicted average genetic COI of 10% or less for all CDR verified litters – our highest benchmark standard. This number is based on a consensus of advice and data from the world of population genetics and we believe that starting with 10% as a goal – and consistently monitoring COI in our dogs – is the best way to plan for long-term health. Here is a summary of the reasoning behind this guideline, and the data that supports it.
Genetic Diversity- the Basics
COI (Coefficient of Inbreeding) = the percentage of an animal’s DNA that is identical because it came from the same ancestor on both sides of the pedigree.
Higher COI means a higher chance that two silent, harmful recessive mutations will “meet up” and be expressed.
Pedigree vs. genomic COI: traditional pedigree math assumes founders were unrelated. Modern DNA tests measure the actual duplicated segments. Genomic COI is therefore almost always higher—and more accurate—than the pedigree estimate.
COI in wild populations of mammals
These are probability bands, not magic lines. A few species persist at slightly higher COIs after “purging” the worst mutations, but that is the exception, not the rule.
COI in Domestic Dogs
A genomic study of 2,000 dogs across 227 breeds found an average COI of ~25 %—already equivalent to offspring of full siblings. Dogs at the higher end had more health problems and shorter lives.
Across both wildlife and dogs, each 10 % rise in COI is associated with roughly a 5–15 % loss in key traits such as litter size, juvenile survival, or lifespan.
Conservation biologists begin to worry once a population’s average COI climbs above ~10 %. Pure-bred dogs, on average, sit well above that red line.
Genomic COI (Embark) vs. heterozygosity (Better Bred)
In many pure-bred circles the go-to number is a “diversity score” or % heterozygosity from a SNP panel; breeders swap stud dogs until that figure creeps up, then claim that the litter is plenty outcrossed.
Some critics even claim that a high heterozygosity score is just as good as—if not better than—checking genomic COI. Trouble is, heterozygosity only tells you how many markers are different, not whether the same long blocks of DNA are duplicated on both chromosomes, so it can miss hidden autozygous stretches that COI picks up.
Embark’s genomic COI measures the share of a dog’s DNA that is identical by descent, but it can’t tell good homozygosity from bad. Better Bred’s heterozygosity score flips the lens—reporting the percentage of non-identical markers—yet critics note that high heterozygosity can be achieved by mixing distant lines without actually purging recessive diseases, so most population geneticists treat COI and heterozygosity as complementary metrics rather than stand-alone health grades.
Why we still need to think about diversity, not just COI
Our first rule is simple: keep every litter’s genomic COI under about ten percent, because high inbreeding is the fastest way to stack health problems. But that number doesn’t tell a breeder everything. If you mostly use dogs from your own , your gene pool can still shrink a little each generation—even with low COI matings—just by chance. Running a quick diversity test (a heterozygosity panel) now and then shows whether you’re hanging on to the rarer genes in your line. We don’t make this test mandatory yet, but we do recommend it for anyone who’s breeding mostly within one family of dogs. Think of COI as your “don’t marry your cousin” rule, and heterozygosity as a reality check that you’re not slowly painting yourself into a genetic corner.
CDP is not against breeding dogs with higher COIs. We’ll still register higher-COI litters as “Breeder’s Choice.” They’re sometimes useful (e.g., to keep a rare trait), but they won’t earn the CDR Verified badge. We are in favor of promoting best practices over the long haul, and we believe that keeping COI at close to10% or lower is a solid, evidence based strategy that breeders can use to help protect the health of our canine friends.
References
Bannasch D. et al. 2021. Canine Medicine & Genetics.
Robinson J.A. et al. 2019. Isle Royale wolves. Science Advances.
Johnson W.E. et al. 2010. Florida panther rescue. Science.
O’Brien S.J. et al. 1985. Cheetah bottleneck study. Science.
Kardos M. et al. 2018. Scandinavian wolves. Nature Ecology & Evolution.
Axelsson E. et al. 2022. Genome-wide association implicates NEBL in myxomatous mitral valve disease in Cavaliers. Frontiers in Genetics.
Birkegård A. C. et al. 2016. Prevalence of mitral valve prolapse in 1,139 asymptomatic CKCS. Journal of Veterinary Internal Medicine.
Hedan B. et al. 2023. Multi-omics mapping of histiocytic sarcoma risk loci in Bernese Mountain Dogs. PLOS Genetics.
Kent M. S. et al. 2018. The Golden Retriever Lifetime Study: establishing an observational cohort. Canine Medicine and Genetics.
Pastor M. et al. 2019. Genomic inbreeding and cancer risk in Golden Retrievers. Scientific Reports.
Stoffel M. et al. 2021. Cross-species inbreeding-depression patterns. Nature Ecology & Evolution.