Linebreeding vs Outcrossing: The 10% F Rule

Hold every planned inbreeding under 10% F and you can linebreed four or five generations before the bill arrives. Push a 2×2 repeat to 12–15% and the loss shows up in kittens before it shows up in type: litters of two instead of five, fading kittens in the first 72 hours, and 4-week weights that sit 10% under the last litter. The working rule is blunt. One 2×2 every second generation, a 3×3 in between, and an outcross the first time a line underperforms twice in a row.

What the F value actually measures

F is the chance that the two copies of a gene a kitten carries at any one location are identical by descent — both inherited from the same ancestor, not merely from the same breed. It is computed from a pedigree, so the depth you feed in decides the answer. On five generations you get one number; add the sixth and it typically climbs 1–3 points. Pick one method, write it in your records, and never compare your figure with a certificate that used the other one.

The 10% ceiling is not a law of cat genetics; it is the level at which livestock and laboratory breeding programs start measuring inbreeding depression in litter size and neonatal survival. Below 10%, most cat lines still produce 3–6 live kittens with birth weights in the normal band. Above 15%, you are running an experiment on your own queen.

Reading a 2×2 and a 3×3

The two numbers say where a shared ancestor sits, not how closely the pair is related. A 2×2 breeding repeats an ancestor in the second generation on both sides — sire and dam share a grandparent. A 3×3 repeats one two generations further back; they share a great-grandparent. When the depths differ, the cross gets lighter: 2×3 means the same cat is a grandparent on one side and a great-grandparent on the other, and the resulting F is roughly half of a 2×2.

Whether one ancestor or two is shared matters just as much. First cousins that share both grandparents land near 6.25%; sharing only one grandparent drops the kittens to about 3.1%. That difference is the whole reason 2×2 is workable: you choose the single ancestor you want to concentrate, not the pair.

Mating	Shared ancestor	F of kittens	How often
Parent × offspring, full sib × full sib	1×1	about 25%	Never as a plan
Half sibs, grandparent × grandchild	2×1	12.5%	Once, at the end of a line
2×2, both grandparents shared	2×2	6.25%	Workhorse cross
2×2, one grandparent shared	2×2	about 3.1%	Safe repeat
3×3, both great-grandparents shared	3×3	about 1.6%	Refreshing type
3×3, one shared	3×3	about 0.8%	Barely linebreeding
Outcross, no common ancestor in 5 generations	—	under 0.5%	When the list below triggers

Where the number misleads you

A kitten at 4% is not automatically safer than one at 9%. F describes the pair in front of you, not the recessives floating through the line. Two linebred parents at 20% each can hand a 4% kitten two copies of the same hidden mutation, because both sides of its pedigree carry it. Ask what the last three generations died of, in writing, before you trust a low percentage.

The second blind spot is the whole breed. If one stud sires 30% of the kittens registered in a decade, the breed-wide effective population shrinks even though every individual pedigree looks clean. Track how many times the same name appears across your own four grandparents, eight great-grandparents, and 16 great-great-grandparents. A name that shows up six times in 16 slots is doing more damage than any single planned cross.

When outcross is not optional

Linebreeding is a tool you keep using until the data tells you to stop. These are stop signals, not preferences:

• Two consecutive litters under three live kittens, or a mean birth weight under 85 g.
• Any confirmed hereditary diagnosis in the line: HCM on echocardiogram, PKD on ultrasound, patellar luxation at grade 2 or worse, progressive retinal atrophy.
• A midline defect, umbilical hernia, or cleft palate in two litters from different queens carrying the same line.
• Stud performance falling off: three matings with no conception, or a semen check under 60% progressive motility.
• Planned F crossing 10% on paper — change the pairing before the mating, not after the litter.

The numbers that tell you it worked

Weigh every kitten on a 1 g kitchen scale at the same hour daily for the first 14 days: birth 90–115 g, a gain of 10–15 g a day in week one, birth weight doubled by day 7–10, tripled by day 21, four times by day 28. A kitten that gains under 7 g a day for two days needs a vet and a bottle, not another week of watching.

Litter size is the slower signal. Average it over five litters from the same queen; a drop from 4.6 to 3.2 is inbreeding depression until proven otherwise. Body size tells the same story at 12 weeks: weigh and measure nose-to-tail-base length, and compare litters raised on the same diet in the same room. An 8–10% drop in 12-week weight across two litters, with no illness in between, is the earliest reliable warning that a line has gone too tight.

Booking the cross without losing a season

Queens cycle seasonally, and a mating placed in the wrong week costs two months. Most breeders aim for the second or third cycle of the year, when follicular activity is stable; the timing detail lives in queen heat cycle timing. Stud side, limit a working male to one litter every 8–10 weeks and give him 48 hours between matings, as laid out in stud cat management. Budget before you commit: a stud fee commonly runs $600–1,500, a planned caesarean $800–2,500, and a lost season is the most expensive item of all. Run the numbers through the cost calculator before the pairing, not after.

A five-generation plan you can write down

1. Generation 1: outcross to the best type you can find. Record F, usually under 1%.
2. Generation 2: 2×2 back to one grandparent. Target F 3–6%; keep every kitten's 12-week weight.
3. Generation 3: outcross again. This is the step breeders skip when they are chasing a look.
4. Generation 4: 3×3 to the generation-1 cat. Target F under 2.5%.
5. Generation 5: decide from data — litter size average, birth weights, echo results — not from the pedigree drawing.

Two of every five planned matings should include an animal with no shared ancestor in five generations. If your plan does not have that ratio, it is not a plan; it is a hope with a spreadsheet attached.

Frequently asked