Tech Tuesdays · No. 1
Her Teeth Say Eight
You can tell a cow's age by pulling her lip and looking at her teeth. Every beef plant in America does it, on every animal, every day. Then the answer is thrown away before it reaches your package — and you eat imported beef instead.
Did you know you can tell a cow's age by looking at her teeth?
Anybody who has worked cattle knows it. You catch her, you pull the lip down, and you count what is in the front of the bottom jaw. There are no top front teeth — a cow has a hard dental pad up there instead — so the whole story is in eight lower incisors, and they arrive on a schedule.
She is born with baby teeth. The permanent ones come in two at a time, from the middle outward, and they take about four years to finish arriving.
The order is the same every time.
The middle pair, the pinchers, erupt at 18 to 24 months and are fully in by two years.
The first intermediates follow at 24 to 30 months, fully developed by three.
The second intermediates at about 36 months, fully developed by four.
The corners, the outside pair, at 42 to 48 months, fully developed by five.
(Mississippi State University Extension, P2779, Estimating Cattle Age Using Dentition)
After that the teeth stop arriving and start wearing.
Past four and a half years you are no longer counting. You are reading. The shape of the wear on the biting surface, how square the tooth has gone, whether the neck of it has started to show.
It gets less certain every year. The University of Florida puts it plainly: past eight years old, telling a cow's age by her teeth is very difficult. (UF/IFAS, AN046)
And the schedule is not the same for every cow.
Mississippi State is direct about this.
Coarse, gritty range forage wears teeth faster than improved pasture. Crossbred cattle keep their incisors longer than straightbreds. Cattle with Brahman blood hold their teeth better than British or Continental breeds.
The guide's own advice is to go and look at cattle of known age in your own country, and adjust the book accordingly.
Which is to say: the method is a national average applied to an individual animal, by eye, by a person, in a hurry.
That is the technology. It is about a hundred and fifty years old and it is still how the United States government decides what your beef is.
Why Anybody Cares
There is a line at thirty months, and it is worth understanding because almost everything else follows from it.
Cattle thirty months and older lose more of the carcass at the plant.
Under federal rule the brain, skull, eyes, trigeminal ganglia, spinal cord, most of the vertebral column and the dorsal root ganglia all have to come out and be destroyed.
Tonsils and a section of small intestine come out of every animal regardless. The rest of that list applies only to the older ones. (9 CFR 310.22)
The regulation also decides the tie. Those materials are deemed to be from an animal thirty months or older unless the plant can prove otherwise. Absent proof, she is old.
So how does the plant prove it? It looks at her teeth.
The federal instruction is specific, and it is the reason this whole piece exists:
Cattle showing one or both of the second pair of permanent incisors above the gum line are considered thirty months or older.
Nought, one or two permanent incisors: under thirty months. Three or more: over. (FSIS Directive 6100.4)
One tooth. That is the whole difference between two commercially different animals, and it is called by a human being on a moving chain.
What It Costs Her Owner
A great deal, and the number is public.
USDA publishes what packers pay over and under the Choice base price every week. The discount for a carcass judged too mature — the trade calls it a hardbone — is not a rounding error.
−$34.49 per hundredweight, weighted average, with a range of −$20.00 to −$55.00.
Deeper than the Select discount at −$21.25.
About level with a dark cutter at −$36.85, which is a genuine quality defect.
(USDA AMS, 5-Area Weekly Premiums and Discounts, LM_CT169, 31 August 2026)
On an eight hundred and fifty pound carcass the average hardbone discount is a little over $290 a head. That multiplication is ours, not the government's.
Behind the discount sits the grading standard itself. Prime, Choice, Select and Standard are restricted to beef from young cattle. Commercial grade is for cattle too mature for those. Utility, Cutter and Canner take animals of any age at all. (USDA AMS, United States Standards for Grades of Carcass Beef)
Read that again, because it is not a statement about how the meat tastes.
It is a rule that an older animal cannot be graded Choice. Not that she is not Choice. That she is not allowed to be.
How Often the Teeth Are Wrong
Often enough to matter, and the government has measured it.
Texas Tech ran the study the USDA needed on this exact question, checking both methods against cattle whose real birth dates were known.
Dentition correctly sorted more than 94% of cattle at the thirty-month line.
5.9% of cattle that genuinely were thirty months or older were called young.
Judging by bone instead of teeth missed 59.7% of them.
(Miller, Texas Tech University, Use of Dentition and Skeletal Maturity to Verify Ages of Cattle Harvested in U.S. Fed Beef Plants, via the USDA National Agricultural Library)
Teeth beat bone by a mile. That finding is why the grading standards were amended in 2017 to let dentition set maturity at all.
The error in the other direction is a young animal wrongly called old and docked for it. As far as we can find, nobody has measured that one against known birth dates.
USDA's own economists modelled it at about one per cent of graded cattle, and put the value of fixing it at as much as $55 million a year to producers. (USDA AMS Agricultural Analytics Division, May 2016)
One per cent sounds small until you notice it is one per cent of twenty-one million carcasses.
This is live right now. USDA has an open request for information on whether to drop the skeletal maturity requirement altogether for carcasses shown to be under thirty months, and the comment period was extended this month — expressly to allow completion of research on dentition. Comments close 7 December 2026. (91 FR 57125, 8 September 2026)
The Part Where a Phone Does It
This is the technology half, and it is why we started a Tuesday series.
A company called MolarMetrics has built a model that reads dentition from a photograph.
You pull the lip the way you always did. You take the picture with a phone. A few seconds later you have an age estimate, a confidence score, and a record attached to the animal.
They sell it by the scan, from about a dollar seventy down to under a dollar in volume. There is an API, so it can sit inside the software a stockyard already runs.
Their stated markets are the four places the answer is worth money: producers making management decisions, stockyards buying and selling, auditors verifying herds for import and export, and plants sorting either side of the thirty-month line.
We are not going to repeat their accuracy figure.
The website leads with ninety-six per cent. That is given as model confidence — the model's own score on its own answer, which is not the same thing as being right ninety-six times in a hundred.
They say a professor is publishing peer-reviewed work on it. When that is out we will read it and report what it says.
But the idea does not depend on their number, and it is the right idea. If a person doing this by eye misses nearly six per cent of old cattle, then a camera that is merely consistent is worth having.
A machine does not get tired at the end of a shift. It does not have a good eye or a bad one. It writes down what it saw, with a date on it, and that record can travel.
Where it travels is the interesting part.
In Spain, Old Is the Expensive One
Here the story turns over, because the American assumption that old means bad is not a fact about beef. It is a fact about America.
Spanish law defines the categories by age outright. A buey is a castrated male over forty-eight months. A vaca is a female over forty-eight months. The words on the label are age categories, and they always have been. (Real Decreto 1698/2003, Anexo I)
Those are not discount categories.
In León, Bodega El Capricho sells ox at €155 a kilo, with a premium loin cut at €192. Its own shop describes the animals as over four years, aged in some cases beyond a hundred and forty days. Older cow steaks from the same house run €42 to €80. (prices from the producer's own shop, checked 14 September 2026)
For comparison, a dry-aged USDA Prime bone-in ribeye from one of the best-known American producers works out near $153 a kilo. The Spanish old-animal product is not a cheap alternative to American Prime. At the top it is priced level with it or above it.
We should be careful here, because a lot gets repeated about Spanish beef that nobody checks. You will read that these oxen are fifteen or seventeen years old. The Spanish outlets that actually interviewed the man who buys them put the range at five to ten years, and one documented animal at nine. We found no paperwork for anything older, so we are not printing the bigger numbers.
And a real buey is rare to the point of near-fiction.
Spain's own agriculture ministry classified over 1.1 million carcasses across forty-two abattoirs. Castrated males — the category every possible buey must fall into — made up 0.27% of them. Cows that had calved were 16%.
There is roughly sixty times more old cow in Spain than old ox, whatever the menu says. (MAPA, Caracterización de la clasificación de canales de vacuno en España)
Here is the part that matters most, and it is the part the romance leaves out.
Age is not what makes that meat. Finishing is.
Galician research on dairy cows put a number on it. A hundred and twenty days on good grazing took carcasses from 247 kilos to 333, lifted fat cover from 5% to 13%, and moved the price from €1.85 to €2.40 a kilo.
The same programme found the gap between a poorly conditioned Friesian cow and a well-finished one was €415 against €1,570. (Centro de Investigacións Agrarias de Mabegondo, via Campo Galego)
Same animal. Same age. Nearly four times the money, because somebody fed her properly for four months before she went.
What Actually Changes in the Meat
Something does change, and it would be dishonest to pretend otherwise.
The best evidence is American and recent.
Texas Tech and Colorado State put grain-finished mature cattle against young cattle at five matched marbling levels. 448 mature carcasses, 423 young.
They ran shear force tests, then sat three hundred and twenty consumers down in sixteen sessions to eat the steaks.
Young beef was more tender in four of the five marbling categories. That is real and it is the honest finding.
But on everything else:
Maturity had limited to no impact on juiciness, flavour intensity, and flavour liking.
And well-marbled mature beef came out close to young beef overall. The authors concluded that grain-finished mature cows gave a similar or slightly reduced eating experience, and that more marbling offsets the tenderness difference.
They went further. They suggested the industry build branded programmes around long-finished mature cattle, to move that product beyond the grinder. (McKinzie et al., Foods 2025, 14(4), 676)
A French study of a hundred and eleven cull cows between fifty and ninety-five months found finishing practice mattered less than the animal herself, and that cows finished on pasture carried more colour in the fat. (Couvreur et al., Foods 2019, 8(4), 141)
That yellow fat, by the way, is not age. It is carotene out of grass, laid down over a long time on pasture. An old cow finished in a pen does not have it. The colour is a record of what she ate and for how long, which is exactly why the Spanish protected-origin rules specify it.
One honest gap. We looked for peer-reviewed work on the eating quality of old Rubia Gallega specifically, the breed most associated with this trade, and could not find any. The Spanish research is about profitability and carcass grading. The sensory science is American and French. When somebody tells you the science proves Galician old cow is better, ask them for the paper.
Now the American Arithmetic
This is where the whole thing stops being a curiosity about teeth.
There are 86.2 million cattle in the United States. It is the smallest herd since 1951. This year's calf crop, at 32.9 million, is the smallest since 1941. (USDA NASS, 30 January 2026)
Ground beef at retail is $6.92 a pound. (U.S. Bureau of Labor Statistics, August 2026)
And it is the cheap end of the animal that is climbing fastest. Wholesale eighty-five per cent lean ground beef is up about 61% since 2022. Tenderloin, over the same stretch, is up about 27%. (Peel, Oklahoma State University Extension)
Ground beef is now roughly half of all the beef Americans eat.
And here is the thing almost nobody outside the trade knows.
You cannot make hamburger out of fat trim alone. You need lean to blend it with, and the leanest beef in the country comes off older cows.
It takes the lean from three or four cull cows to absorb the fat trim from one fed steer.
America does not have enough old cows.
6.1 billion pounds of beef imported in 2026, up 14% on the year.
About 68% of it is lean trimmings and lean beef for grinding.
Roughly 39% of all the lean going into American ground beef now comes from abroad.
Australia, Brazil, Canada, Mexico, New Zealand.
(USDA Economic Research Service, Livestock Dairy and Poultry Outlook, August 2026; Peel, Oklahoma State University Extension, 14 September 2026)
The long-run average for imported trim was about a quarter. It is now closer to four pounds in ten.
So the American cow that does go to slaughter is not being wasted, and she is not cheap.
Ninety per cent lean cow trim was $435.21 a hundredweight on 11 September. The entire Choice boxed beef cutout, that same week, was $375.94. (USDA AMS, week ending 11 September 2026)
The lean off an old cow is worth more per pound than the average pound of Choice beef in America.
She is docked $34 a hundredweight on the rail for her teeth, and her trim then sells above the Choice cutout. Both of those things are true at the same time, and the gap between them is a system that has not caught up with its own market.
The Thing That Is Already Decided For You
Every carcass in this country gets its age determined. By teeth, at the plant, on the day.
That number decides whether she can be Choice. It decides what her owner is paid. It decides which parts of her are destroyed.
And then it stops. It never reaches the package.
There is no age on a pound of ground beef, and no federal standard for putting one there. An age claim is not defined anywhere in the meat inspection regulations. It would be a special claim, approved case by case, on documentation the producer would have to build from scratch.
Europe already does this, in one direction. EU rules require the animal's age category to appear on the label for cattle twelve months and under. The machinery exists. It has been running since 2008. It is simply pointed at the young end. (EU Regulation 1308/2013, Annex VII)
A handful of American restaurants have gone the other way on their own. There are eight and ten year old animals on menus in Las Vegas and Boulder and Chicago, with the age printed next to the price, and the price is high. The American company selling Spanish-style old cow to households imports its animals from Tasmania.
From Tasmania. To a country with five million cull cows a year.
What We Are Not Saying
We are not saying old beef is as tender as young beef. The best study on it says clearly that it is not, at four marbling levels out of five, and we are not going to pretend a consumer panel said something it did not.
We are not saying there is a pool of cheap old beef sitting unwanted. We thought there might be when we started. There is not.
Lean cow trim is one of the dearest things in the American meat complex right now. Anyone telling a rancher his cull cows are being thrown away has not looked at a price sheet.
We are not saying anybody is doing anything crooked. The thirty-month rule exists for a reason and that reason was a disease that killed people. The grading standards were written when the country had cattle to spare.
What we are saying is narrower. The country is short of exactly the animal its own grading system is built to discount, and it is buying the shortfall from five other countries. The one fact that would let a customer choose differently is already collected on every carcass, and then thrown away. A camera and a model that reads teeth makes it portable for about a dollar. Whether it ever reaches a label is a decision, not a technical problem.
So
You bought ground beef this month.
Nearly half the beef eaten in America is ground. About four pounds in every ten of the lean in it crossed an ocean or a border to get here.
Nobody asked you about that. It is not on the package. There was no choice offered and no discount for taking it.
The question we would put to you is the one the label never asks.
Would you rather have meat from an older American cow, raised on grass a few states away by somebody whose name you could learn — or imported beef from a country you know nothing about?
You are already answering it. You are just not the one answering.
Sources
Mississippi State University Extension, P2779, Estimating Cattle Age Using Dentition — incisor eruption schedule and the effect of diet, breed and geography on it.
University of Florida IFAS, AN046 — eruption and wear, and the limits of the method past eight years.
USDA FSIS Directive 6100.4 — the thirty-month dentition threshold and inspector instructions.
9 CFR 310.22 — specified risk materials, what is removed at thirty months and older, and the presumption against the establishment.
USDA AMS, United States Standards for Grades of Carcass Beef — maturity groups A to E and the restriction of Prime, Choice, Select and Standard to young cattle.
USDA AMS, 5-Area Weekly Weighted Average Direct Slaughter Cattle Premiums and Discounts, LM_CT169, 31 August 2026 — the hardbone discount.
Miller, Texas Tech University, Use of Dentition and Skeletal Maturity to Verify Ages of Cattle Harvested in U.S. Fed Beef Plants, via the USDA National Agricultural Library — 94% and 5.9% for dentition, 59.7% for skeletal maturity.
USDA AMS Agricultural Analytics Division, Economic Assessment of the Request to Modernize the U.S. Standards for Grades of Carcass Beef, May 2016 — the false-old estimate and the $55 million figure.
91 FR 57125, 8 September 2026 — extension of the comment period on the carcass beef grading standards to 7 December 2026.
Real Decreto 1698/2003, Anexo I (BOE) — Spanish sales denominations by sex and age, including buey and vaca at over forty-eight months.
MAPA, Caracterización de la clasificación de canales de vacuno en España — carcass category shares across 1,173,109 carcasses.
Centro de Investigacións Agrarias de Mabegondo, via Campo Galego — the effect of finishing on cull cow carcass weight, fat cover and price.
McKinzie, Garmyn, McKinzie, Koohmaraie, Legako, Woerner and Miller, Foods 2025, 14(4), 676 — consumer panel comparing grain-finished mature and young beef at matched marbling.
Couvreur, Le Bec, Micol and Picard, Foods 2019, 8(4), 141 — cull cow characteristics, finishing practice and meat quality.
USDA NASS, Cattle report, 30 January 2026 — 86.2 million head and the 32.9 million calf crop.
U.S. Bureau of Labor Statistics, average price of ground beef, August 2026.
USDA Economic Research Service, Livestock Dairy and Poultry Outlook, August 2026 — import volumes and forecast.
Derrell Peel, Oklahoma State University Extension, 14 September 2026 — the share of imported lean in American ground beef and the lean-to-fat blend arithmetic.
USDA AMS, National Weekly Boxed Beef Cutout and cow cutout reports, week ending 11 September 2026.
EU Regulation 1308/2013, Annex VII — mandatory age category labelling for cattle twelve months and under.