What a COI number actually means
for your clutch
Every breeder has heard that inbreeding is bad. Very few can say what a coefficient of 6.25% actually predicts, or why 12.5% is the line where experienced breeders stop arguing and find another bird.
- COI is the probability that a chick inherits two identical copies of the same gene from a shared ancestor
- Below 6.25% is ordinary. Above 12.5% has measurable effects on hatch rate and chick survival
- The damage is cumulative across generations, not within a single pairing
- Most real inbreeding hides one generation back, in the grandparents nobody wrote down
What the number is actually measuring
The coefficient of inbreeding, written F, is the probability that a chick inherits two identical copies of the same gene, both descended from one shared ancestor. It is not a score of how "inbred" a bird looks. It is a straightforward probability, and it is calculated from the pedigree alone.
Sewall Wright published the method in 1922 and it has not needed replacing. For each ancestor that appears on both sides of a pairing, you count the generations back to that ancestor from the sire and from the dam, and add up the contributions. Two birds that share nothing contribute nothing. A bird that appears on both sides twice contributes twice.
COI does not tell you a pairing will go wrong. It tells you how much of the chick's genome is likely to be doubled up, and doubled-up genes are where recessive problems become visible.
Why the numbers land where they do
The values are not arbitrary thresholds somebody voted on. They fall out of the arithmetic, and every one of them is an exact binary fraction.
| Pairing | COI | What is shared |
|---|---|---|
| Full siblings | 25% | Both parents |
| Parent to offspring | 25% | Direct line |
| Half siblings | 12.5% | One parent |
| Aunt or uncle to niece | 12.5% | A parent's full sibling |
| First cousins | 6.25% | Both grandparents |
| Half first cousins | 3.125% | One grandparent |
| Second cousins | 1.5625% | A great-grandparent |
Wright's coefficient across three generations, assuming the shared ancestors are not themselves inbred. If they are, every figure rises.
Notice that each step back through the pedigree quarters the contribution. That is why a shared grandparent matters so much less than a shared parent, and why a shared great-grandparent barely registers at all.
What actually happens to the clutch
Inbreeding depression is the general term for the effects, and in birds it shows up in a predictable order. Fertility drops first, then hatch rate, then chick survival in the first two weeks, then adult size and vigour. You usually notice it as clear eggs and dead-in-shell long before you notice it as a visibly weak bird.
This ordering matters practically. A breeder who watches only the fledged chicks will conclude nothing is wrong while the real loss is happening at candling. If your hatch rate is falling and your pairs are otherwise healthy and well fed, check the pedigrees before you change the diet.
The effect is also not a cliff. It is a slope. A single pairing at 6% is not meaningfully different from one at 5%. The difference between 6% and 25% is enormous.
Why line breeding is not the same
Line breeding is deliberate, low-level inbreeding used to fix a mutation or a trait in a line. It is a legitimate technique and most established mutations exist because somebody did it on purpose.
The distinction is not the number in a single generation. It is whether the number repeats. One pairing at 6% costs almost nothing. The same pairing repeated for four consecutive generations compounds, because each generation inherits the doubled-up genes of the last and adds its own.
Treat COI as a running total across your line, not as a per-pairing pass or fail. If you use a pairing in the caution band, plan the next generation out to unrelated stock before you commit to it.
The generation everyone forgets
Two birds with completely different parents can still share a grandparent, and that pairing carries 3.125% per shared grandparent. Without the grandparent layer you would read it as unrelated and record a zero.
This is the single most common way real inbreeding enters an aviary. Nobody deliberately pairs siblings. Plenty of people unknowingly pair the grandchildren of one popular cock bird, because that cock bird was good and everybody used him.
It is also why a pedigree that stops at the parents is close to useless for this purpose. Three generations is the practical minimum, which is seven birds: the chick, two parents, four grandparents.
Free COI calculator. Wright's method across three generations, in your browser. No account, nothing stored.
What the number will not tell you
COI is blind to everything outside the pedigree you gave it. It cannot see that both grandparents came from the same small local population, so the "unrelated" birds are quietly cousins. It cannot see that one line carries a fault the other does not. It cannot see nutrition, humidity, or a hen that has been laying without a rest.
It also cannot see missing records. A pedigree with four blank grandparents will return a comfortable zero, and that zero means "no shared ancestors in the data I have", not "these birds are unrelated". Incomplete pedigrees flatter every pairing.
Use it as one input. A 3% pairing between two birds with a known weakness is a worse idea than an 8% pairing between two robust ones.
Common questions
What is a safe COI for lovebirds?
Below 6.25% is generally considered safe and covers most unrelated or distantly related birds. Between 6.25% and 12.5% is workable once, typically to fix a mutation, but should not be repeated in the following generation. Above 12.5% carries measurable risk to hatch rate and chick survival.
Does a high COI mean the chicks will be deformed?
No. It raises the probability that recessive faults already present in the line become visible, and it depresses fertility, hatch rate and early survival. Most high-COI clutches produce normal-looking birds. The cost usually shows up as fewer chicks, not obviously damaged ones.
Is line breeding the same as inbreeding?
Line breeding is deliberate inbreeding at a low coefficient, used to fix a trait. The mechanism is identical. The difference is intent and, more importantly, whether it is repeated. The risk accumulates across generations rather than within one pairing.
Why does the calculator ask for grandparents?
Because most real inbreeding hides one generation back. Two birds with different parents can share a grandparent, which produces 3.125% per shared grandparent. Without that layer you would record the pairing as unrelated.
Does this work for budgerigars and other species?
Yes. Wright's coefficient is species-agnostic and the arithmetic is identical for budgerigars, cockatiels and finches. Only the tolerance thresholds differ slightly between species.