A landrace is a crop population that farmers have grown and reselected for generations without it ever passing through a formal breeding program. The U.N.'s Food and Agriculture Organization publishes the working definition most crop scientists use: a "dynamic population of a cultivated plant that has historical origin, distinct identity and lacks formal crop improvement, as well as often being genetically diverse, locally adapted and associated with traditional farming systems." Coffee has one of the largest surviving populations of this kind in modern agriculture, and almost all of it grows in one country: Ethiopia, the place arabica coffee came from in the first place.
What Actually Makes a Coffee Count as One?
Plant scientist Mario Negri's fuller version, in the same FAO resource book, adds the piece that matters most to a coffee buyer: it "lacks 'formal' crop improvement, is characterized by a specific adaptation to the environmental conditions of the area of cultivation... and is closely associated with the traditional uses, knowledge, habits, dialects, and celebrations of the people who developed and continue to grow it." No laboratory selected it. No breeder crossed two parents to fix a trait. Farmers kept planting seed from the trees that did well on their own hillside, year after year, and the population adapted to that hillside instead of to a spec sheet.
That is a real agricultural category, not a marketing word, and it sits between two things it is often confused with. It is not the same as a named, bred variety like SL-28, which a government laboratory in Kenya selected out of a specific parent population in 1935 and can date to the year, the place and the goal. And it is not the same as a truly wild plant that nobody farms at all. What FAO is describing is a working crop, cultivated, harvested and sold for income, that simply never passed through a breeder's hands.
Coffee is not unusual in having this problem. Most of the world's major food crops went through the same split at some point in the last century: wheat, maize, rice and potatoes all had enormous regional pools of farmer-kept populations before national breeding programs replaced most of them with a shorter list of higher-yielding, disease-resistant lines. What makes coffee's case unusual is timing and geography. The replacement happened later in coffee than in grain crops, and the deepest pool of unreplaced genetics sits almost entirely inside one country's borders, which is a much smaller and more fragile place to keep a whole species' genetic history than a scattering of farms across a continent.
Is It the Same Thing as Wild Coffee?
No, and the industry's own shorthand blurs this more than it should. Coffee writing regularly calls Ethiopia's unbred trees "wild," and the confusion is understandable, because the population's genetics trace back to truly wild forest coffee with no farmer involvement at all. But most of what actually gets picked and sold grows in a farmer's forest garden, semi-forest plot, or backyard, tended, harvested and sold for cash by a family that depends on it every year. A plant can be genetically close to wild and still be a working crop that somebody plants, weeds and picks on purpose. Those are different facts, and a bag that calls a coffee "wild" when it means something a farmer manages is overstating how little human hands actually touched it.
The reverse error is just as common: treating the word as a synonym for "unimproved" in a way that implies lower quality. It does not mean that. It means unselected for a specific trait by a laboratory, not unselected for excellence by anyone. A farmer who has kept the best-tasting trees on his own hillside for three generations has been running a breeding program of his own, informally, often for longer than any named variety has existed. The selection pressure is real; it is just applied by a household deciding which trees to keep replanting rather than by a research station running a controlled trial. Genuine Origin, one of the importers we buy through, covers the same ground from the trading side: a lot arriving with no variety name on the paperwork is not a defect in the record, it is an accurate description of what is actually in the bag.
How Much of Ethiopia's Crop Actually Grows This Way?
Far more than most drinkers assume. A 2023 coffee-and-biodiversity policy brief covering southwest Ethiopia's biosphere reserve breaks the country's crop down by how it is grown: forest coffee, picked from trees growing inside the natural forest canopy with no other management, accounts for 10 percent of national production. Semi-forest coffee, grown in a cleared and thinned forest understory in the buffer zone around the reserve, accounts for 35 percent. Garden coffee, the same trees moved into a farmer's own backyard and intercropped with food, is 50 percent. Only plantation coffee, grown as a monoculture on commercial or state-run land using, in the brief's own words, "high-yielder seedlings and disease resistant varieties," makes up the remaining 5 percent of what the country produces in a given year.
Add the first three together and roughly 95 percent of everything Ethiopia grows traces to unselected, locally adapted populations rather than a laboratory. The 5 percent that does not is also the only 5 percent planted on purpose with a named, bred variety. That ratio is close to the reverse of what most origin countries look like, where a handful of laboratory selections cover almost the whole crop and a farmer growing something off-catalog is the exception rather than the rule, not the other way around. Kenya, where our own Kiambu lot grows, sits near that other end: a small government research program did the selecting decades ago, and most of the country's farmers plant from that short list today.
Why Does the Flavor List Run So Long?
Because the bag is telling the truth about what is actually in it. A single named variety like SL-28 is, by definition, one genetic line: distinct, uniform and stable, the three conditions the International Union for the Protection of New Varieties of Plants uses to define a variety at all, and which World Coffee Research applies to its own catalog. An Ethiopian garden lot fails that test on purpose. One hillside can hold dozens of genetically distinct trees, never separated at harvest, never separated at the mill, so the cup carries the sum of all of them rather than the signature of one. That is why a tasting-note list on this kind of Ethiopian lot routinely runs longer than the list on a single-variety coffee from almost anywhere else. It is a fact about the plant population in the cherry, not a fact about how hard a roaster is trying to describe it, and it is also why two bags from the same washing station in different years can read differently: the mix of trees behind the blend shifts from harvest to harvest in a way a single cloned variety never does.
Why Do Plant Breeders Still Go Looking for It?
Because it is the raw material they cannot manufacture. A breeding program can cross two known parents and select for a trait in a few generations, but it cannot invent new genetic variation from nothing; it can only recombine what already exists somewhere in the crop's gene pool. Ethiopia's gardens and forests are the deepest reserve of that variation left in Coffea arabica, because arabica as a species passed through a documented genetic bottleneck when it left Ethiopia centuries ago, and almost everything planted commercially outside Ethiopia today descends from a small slice of what was available at home. When leaf rust, a new pest, or a warmer growing region shows up somewhere a named variety cannot handle it, the trait that solves the problem is more likely sitting in an uncatalogued Ethiopian tree than in anything already on a breeder's shelf.
El Salvador Finca San Luis is a real, current example of that pipeline, not a history lesson. The coffee grows Centroamericano H1, a first-generation hybrid World Coffee Research bred by crossing a rust-resistant parent with Rume Sudan, a population collected in Ethiopia decades earlier for exactly this kind of work, which our own longer piece on variety names traces in full. The hybrid itself is bred and deliberate, and has to be replanted from a nursery every generation rather than from its own seed. But half its genetics came from an unselected Ethiopian tree that somebody walked into a garden to find, years before leaf rust or a shifting climate made that cross worth making. The raw material does not stop being useful once a breeding program picks it up. It becomes the source for the next generation of named varieties, which is the main reason plant scientists keep arguing for protecting what is left of it rather than letting it get replaced field by field with higher-yielding clones that all share the same handful of parents.
What Does This Mean If You're Choosing Between Bags?
Treat the two kinds of coffee as answering different questions rather than as better and worse. A named variety tells you what to expect: the same cross, planted on purpose, should taste recognizably like itself from one harvest to the next, which is why a cafe building a house blend or a wholesale account standardizing a menu usually wants one. An unselected Ethiopian lot tells you something else entirely: this is what one farmer's hillside produced this year, assembled from whatever was already growing there, and it will not taste quite the same next year even from the same address. Neither answer is the honest one and the other the marketing one. They are simply describing two different relationships between a plant and the person who grew it, and our own longer piece on Ethiopian coffee specifically goes further into what that unpredictability actually tastes like in the cup.
Does Our Kenya or El Salvador Count?
No, and they miss it in two different ways worth knowing. Kenya Kiambu, linked above, lists two varieties, SL-28 and SL-34, both selected at a government laboratory outside Nairobi in the 1930s and both passing the distinct-uniform-stable test on their own, which our own varietal piece works through in detail. That is the opposite case: two known parents, a known date, a known goal. Finca San Luis is the case just above it: bred, but built in part from the raw material this whole article is about. Between the two of them, our current shelf runs entirely on genetics somebody selected on purpose, at some point, for some reason, which is worth knowing if you are choosing between them for what you actually want in the cup.
Do We Have One on the Shelf Right Now?
No, and it is worth saying plainly rather than leaving the shelf to answer the question on its own. We have carried two of these Ethiopian lots this year, Buku Abel and Dry Process Hambela Buku, and both sold through every size we had before this was written. That is a supply fact, not a quality signal. This kind of coffee tends to move through a small roaster's shelf faster than a laboratory variety does, because there is only ever one farmer's hillside behind each lot rather than a whole region planted to the same clone, and once it sells out there is no second identical lot to reorder from the same importer at the same price. We will carry another Ethiopian lot like it when the harvest allows. Until then, our shelf runs on named varieties, which is a real trade-off and not a downgrade: more consistency, less of the genetic noise that makes this kind of coffee taste the way it does, and a cup you can expect to taste the same way next time you order it. If you want the fuller chain behind where any of our coffee actually comes from, including how it clears customs and reaches Pensacola, our origins and sourcing writing covers that ground separately.
Key takeaways
- A landrace is a crop population that farmers reselected for generations without a formal breeding program, adapted to its own ground rather than bred to a spec.
- It is not the same as wild coffee, which no farmer plants or sells, and not the same as a named variety like SL-28, which traces to a known breeder and date.
- Roughly 95 percent of Ethiopia's coffee grows this way, across forest, semi-forest and garden systems; only the remaining 5 percent, grown as plantation monoculture, is a bred variety.
- Breeders still collect it to build new hybrids: Centroamericano H1, grown at Finca San Luis, has an Ethiopian parent population, Rume Sudan, in its ancestry.
Inheritance Coffee, Pensacola. Our two Ethiopian lots from this kind of population are both sold through for now. El Salvador Finca San Luis carries a variety bred in part from one, and it is on the shelf today. Kenya Kiambu, two named Kenyan selections from the 1930s, is the other end of the same story. We roast to order every Wednesday and grind either to your brewer at no charge.
