The Essentials of Breeding Records, EPDs & EBVs

Breeding records and a genetic number are not the same thing, and the difference is where most of the confusion lives. A breeding record is yours — the dates, the groups, the weights you wrote down on your own animals. The number, an Expected Progeny Difference (EPD) in cattle or an Estimated Breeding Value (EBV) in sheep, is something a breed association computes from your data and sends back to you. Confuse the two and you expect a record to do a job only an evaluation can do — or expect an evaluation to fix a gap your records never filled.

This is the canonical guide to breeding and EPD/EBV records: what to keep, how the genetic numbers are actually produced, and the honest line between a record you own and a prediction you're given. It is about the records, not the mating advice — which males and females to breed stays your call and your vet's.

The record you own vs. the number you're given

The cleanest way to think about breeding records is to separate two layers. The first is the source record — what happened, on your place, written down at the time. The bull went in with the red-tag group on June 1st and came out on August 1st. Cow 1417 weaned a calf at 525 pounds. Ewe 184 lambed twins unassisted. These are facts you observed. They are yours, and they are the raw material of everything that follows.

The second layer is the derived prediction — the EPD or EBV a breed association publishes about an animal. That number is not a fact you recorded; it is a statistical estimate the association's genetic evaluation produced by combining your animal's performance with pedigree and performance data from thousands of related animals across many herds. You submit the source records; the association returns the prediction.

Keeping those layers straight answers a question producers ask every season: can you just figure out your own EPDs from your records? No — not from your records alone, and the reason is structural, not a software limitation.

What to keep per animal

The source record is the layer you control, so it earns the attention. What to keep, and why each field earns its place:

Extension recordkeeping guidance is explicit on the linchpin: record the bull or ram turn-in and removal dates, and any AI dates, for each breeding group. That single habit is what makes next spring's calving watch list and the following year's keep/cull decisions answerable instead of guessed. The pre-breeding herd-health visit is where the soundness exam and the female list get made; this guide covers the exposure and performance records that follow.

EPD/EBV is a derived value

Here is how the number actually gets produced. A breed association runs a National Cattle Evaluation (for cattle) or an equivalent genetic evaluation (GenOvis in Canada, the National Sheep Improvement Program — NSIP — in the US, for sheep). The evaluation takes in performance records its members submit — birth, weaning, and yearling weights, calving ease, carcass data — along with pedigree and, increasingly, genomic (DNA) information. It then estimates how an animal's future progeny are expected to perform relative to the breed average.

The key terms: an EPD predicts the progeny difference — how an animal's calves or lambs are expected to compare. An EBV estimates the animal's own breeding value, and works out to roughly twice an EPD for the same trait (because a parent passes on half its breeding value). Same underlying genetics, two conventions — cattle breed associations tend to publish EPDs, sheep evaluations tend to publish EBVs.

Two things follow from how the number is built. First, accuracy climbs with progeny. A young animal's EPD is an interim estimate drawn from its pedigree; as its own offspring get recorded, the accuracy rises and the number moves. That is why genomically enhanced EPDs exist — DNA data stands in for early progeny and lifts accuracy sooner. Second, the comparison needs many herds. An EBV is calculated by linking related animals across flocks and herds, not by looking at one flock alone; without that connectedness, there is no across-herd comparison to be relative to.

The honest boundary

That second point is the whole reason the misconception dies hard. An EPD or EBV is a relative prediction. It depends on contemporary-group data from many operations, a pedigree structure linking them, and a genetic-evaluation engine that ties it together. None of that lives inside a single producer's records — no app, spreadsheet, or assistant can compute a real EPD from one herd's data alone, because the comparison that makes the number meaningful requires herds the producer doesn't own.

So the honest division of labor is this: you keep the source records — the exposure dates, the actual weights, the calving ease, the parentage. You submit them to your breed association. The association runs the evaluation and sends back the EPDs and EBVs. A system can store both your source records and the numbers that come back, and keep them tied to each animal. What it cannot do is manufacture the prediction — that part belongs to the evaluation.

This is worth saying plainly because the question keeps getting asked: "what's this calf's EPD?" is answerable from a catalog or an association lookup, not from the calf's own file. The file holds the source data the evaluation ran on. The number came from somewhere bigger than your place.

Sale-time documentation

The same source records that feed a genetic evaluation are what a buyer of breeding stock expects to see. "Full history" at a sale means the birth and weaning data, the breeding exposure, the health and treatment records, and — for registered animals — the association's EPDs or EBVs and the registration papers that back them. For registered and seedstock operations specifically, that documentation is the whole product, covered in depth in the seedstock record-keeping guide.

For commercial operations, the bar is the source layer: which bull was with which group, when, and how the calves performed. That is enough to compare sires within your own herd and to stand behind an animal you sell — and it is the same record, kept less formally, that a registered operation submits to an association.

Where Ranch.Bot fits

Ranch.Bot keeps the source records — the breeding exposure by group, the sire, the actual weights, the calving and lambing ease — entered in plain language and tied to each animal, so when preg-check, sale, or the next breeding season comes, you pull an animal's full history instead of digging through notebooks. You review every record before it saves. Ranch.Bot does not compute EPDs or EBVs, and it does not build pedigrees — those come from your breed association's evaluation, and the numbers it sends back are yours to store alongside the records that produced them. What Ranch.Bot handles is the part you own: the source record, kept tied to the right animal, searchable when the decision comes.

Keep the source record clean. The number that comes back is only ever as good as what you sent in.

---

Ranch.Bot turns plain-language notes like "bull 42 in with the red-tag group June 1st, out August 1st" into structured breeding records you can search by animal and group — and you review every record before it saves. Start a 14-day free trial at ranch.bot.