How Much Land Do You Need For A Small Farm?

You want to know how many acres it takes. Fair question, and every article on the internet will hand you a number. Two acres per cow. A quarter acre for twenty chickens. Five acres to feed a family.

Here's the problem with all of them. They're national averages of a quantity that has no meaningful national average.

Texas Parks and Wildlife publishes stocking rates for two parts of the same state. On tame pasture in the Post Oak Savannah, east of Austin, they put it at three to six acres per animal unit. In the Trans-Pecos, out past Fort Stockton, they put it at seventy-five to over two hundred. Same agency, same state, same cow. The spread is somewhere around fifty to one.

So if you're comparing a parcel in Arkansas against one in Nevada, and you're working from a single number, you are going to be badly wrong about one of them. This article is about how to get the right number for the ground you're actually looking at.

The number you want is a function of rainfall

Grass grows where water falls. How much livestock a piece of ground carries comes down to how much forage it produces in a year, and forage production tracks rainfall and soil more than anything else you can control. That's the whole mechanism, and once you've got it you can evaluate any parcel in any state.

Here's what the extension services and agencies actually publish for the ten states we sell in. Read it as a map of climate, because that's what it is.

A few notes on that table, because I'd rather you know where the numbers are soft. The Nevada figure is my arithmetic, not BLM's: they publish about 43 million acres of rangeland authorized for grazing and about 2 million AUMs permitted, and the division is mine. New Mexico State publishes no state figure at all, only a worked example that lands near 62 acres per animal unit after adjusting for water distance and slope. The Arizona publication dates to 2004, which is old but the underlying range science hasn't moved much.

And notice what's missing. Four land-grant universities covering the arid West, New Mexico State, Colorado State, Arizona and Texas A&M, publish methodology and adjustment factors but refuse to publish a single state number. Texas A&M says why, plainly: too often a blanket stocking rate gets used for a whole county or region when it should be tailored to each grazing unit. When four universities that study this for a living won't give you one number, that should tell you something about the articles that will.

The unit everybody uses, explained once

That conversion is the useful part. Five sheep eat about what one cow eats. Six to eight goats, depending on whose table you use. So you don't need a separate acreage figure per species, you need one stocking rate for your ground and then you divide.

Which also means the quarter-acre-per-sheep number you'll see quoted is only true where the cow number is around one and a quarter acres. On good managed pasture in North Carolina, fine. On Arizona rangeland a ewe needs ten to thirty acres, and in the Trans-Pecos more than that.

How to get the real number for your parcel

This is the part worth the whole article, and almost nobody tells land buyers about it.

The NRCS Web Soil Survey will give you forage production for the actual dirt you're looking at. You draw your area of interest on the map, and it returns the ecological site description with potential productivity in animal unit months per acre, or tons of hay per acre, for the soils under that boundary. It's free, it's federal, and it's parcel-specific rather than county-average.

Oregon State Extension points landowners at exactly this for the same reason. Pull the soil survey for the parcel, take the AUM-per-acre figure, and you have a starting stocking rate that belongs to your ground instead of to the internet. Then knock it down for the parts you can't graze: distance to water, slope, timber. New Mexico State's worked example loses almost 40 percent of its theoretical capacity to exactly those two adjustments.

One rule from Colorado State that's worth carrying: plan on using only 25 to 30 percent of the total forage a pasture grows. The rest keeps the plants alive. Stocking at the theoretical maximum is how the Trans-Pecos got wrecked in the 1880s, when ranchers ran four to twelve acres per animal unit on ground that now takes seventy-five to two hundred.

The things that aren't measured in acres at all

Part of what makes these articles confusing is that they put everything in acres, including things nobody measures in acres.

Chickens

Penn State Extension publishes 1.5 to 2 square feet per bird inside the coop, plus 8 to 10 square feet per bird in the outside run. Twenty laying hens works out to roughly 240 square feet total, which is a twelve by twenty footprint. The quarter acre you'll see quoted is about forty-five times more land than twenty chickens need. Chickens are a fencing and predator decision, not an acreage decision.

The vegetable garden

NC State Extension puts a family of four at about 1,050 square feet, and attaches a qualifier that matters: that's for the growing season. If you're preserving for winter, roughly double it. Season length is the biggest variable, which is why the same garden feeds a Florida family nearly year round and a Minnesota family about five months. Their other advice is worth repeating: start smaller than you think and expand when you know you'll keep up with it.

Sheep and goats

Covered above. A mature ewe is 0.20 animal units and a mature goat is 0.15, so take whatever your ground carries per cow and divide. There is no fixed acreage per sheep any more than there is per cow.

Dairy

Depends almost entirely on how much grain you feed. UF/IFAS puts intensively managed Florida pasture at about half an acre per milk cow, with three to four cows per acre in summer, and that's assuming twenty to twenty-five pounds of concentrate a day. Take the grain away and the land requirement climbs fast. There is no national dairy acreage figure and I doubt one is possible.

What the land actually costs

The old version of this article said raw land in rural areas is generally more affordable, which is true and useless. USDA publishes the real numbers every year, by state, for free. Here are the 2026 figures for the ten states we sell in.

Look at the pasture column against the stocking rate table further up. They're the same ranking. New Mexico pasture at $650 an acre and Florida pasture at $7,650 is a twelve-to-one spread, and it lines up almost exactly with how many acres it takes to feed a cow in each place. The market is pricing carrying capacity, and it is doing it accurately. That's two completely unrelated federal datasets telling you the same thing, which is about as much confirmation as you get in this business.

One honest caveat. USDA is valuing farm real estate, which is not the same thing as a five to forty acre recreational or homestead parcel. Small parcels usually trade at a premium per acre. Use these for comparing states, not as a price quote.

Keep reading

Land InvestmentBest States to Buy Farmland: Where Should You Invest?Ten states ranked by what USDA says an acre actually costs, cheapest first. Montana is the cheapest cropland in the country and Iowa is nearly eight times it. Three states that get called affordable are anything but.

How much time it takes, which nobody budgets for

Acreage is the question people ask. Hours is the question that decides whether the place works out.

There's a figure that circulates saying 252 hours a year per acre of vegetables, attributed to MIT. I went looking for that source and could not find it. I'm not going to repeat a number whose origin I can't locate.

What is published is Penn State Extension's crop budgets, and the spread in them is the actual answer. Wholesale tomatoes run about 19 labor hours an acre. Sweet corn is 42 to 52. Strawberries are 119 to 131 operator hours before you pay somebody to pick them. Fresh-market bell peppers run 167 to 230, most of it harvest. That's a thirteen-to-one spread inside one state from one publisher, and the variable is whether you pick by hand.

A diversified market garden, worked by hand, sits at the top of that range and above it. If you're holding down a job, that's the number that will decide the size of your garden, not the acreage.

What USDA actually means by a small farm

Not acres. Income.

USDA's Economic Research Service sorts family farms by gross cash farm income, and a small family farm is one under $350,000. By that definition 86 percent of US farms are small family farms. They operate 40 percent of the agricultural land and account for 17 percent of the value of production. Within that bucket, 39 percent are off-farm-occupation farms, meaning somebody has a job in town, and about 12 percent are retirement farms.

There's no acreage definition anywhere in the federal typology. What there is, and what's more useful to you, is the census count by size. Of the 1.9 million farms in the 2022 Census of Agriculture, 12 percent are one to nine acres and another 30 percent are ten to forty-nine. Forty-two percent of American farms are under fifty acres. If you're looking at a twenty-acre parcel wondering whether that's a real farm, the federal government already counted you.

Self-sufficiency, and the number nobody can give you

I looked hard for a published minimum acreage to feed a family and there isn't one. Not from USDA, not from NRCS, not from any extension service. That's not a gap in the research. It's because the answer depends on diet, climate, season length, soil, water, preservation capacity, skill and how much crop failure you can absorb, and those swing it by more than an order of magnitude. No land-grant is going to publish a number it would have to defend.

The closest honest anchor is a 2016 study in Elementa that modeled US carrying capacity across ten diet scenarios. It found annual land requirements of 0.13 to 1.08 hectares per person, which is roughly a third of an acre to two and a half acres, depending entirely on what people eat. For a family of four that's somewhere between 1.3 and 10.7 acres. Two caveats that matter: that assumes commercial-scale productivity, and it counts grazing land, not just garden. A household growing food by hand will not hit commercial yields.

My own read, from watching buyers actually try this: on a small parcel, water and skill run out before acreage does. People who make it work start with a garden and a few chickens on two acres and grow into the rest. People who don't buy forty acres first and discover that the limiting factor was the hours in their week.

So how many acres?

If you want the version on one line: five to ten acres is a real homestead almost anywhere east of the hundredth meridian, and the same five to ten acres in Nevada is a camp with a garden on it. Both are worth owning. They are not the same purchase.

The way to know is to stop asking the internet and pull the soil survey for the actual parcel. It takes about ten minutes, it's free, and it will tell you more than this article can. Then match the plan to what the ground will carry, rather than buying ground to fit a plan you picked before you knew the rainfall.