Ethiopia

Why this origin matters

Ethiopia is the one large producing country where arabica coffee is a native plant. The species still grows wild in the forests of the south-western highlands, and the genetic evidence, summarised below, places the origin of cultivated arabica in Ethiopia or close to it without being able to say where. It is also Africa's largest coffee producer, growing arabica only, and on the USDA's estimates it trades fourth and fifth place in the world with Indonesia: ahead in 2024/25, a little behind in 2025/26, and forecast ahead again in 2026/27. Two things set its coffee economy apart. Nearly all of it is grown by smallholders, some of it in forest that is managed little or not at all. And Ethiopians drink a large part of the crop themselves, as the consumption figures below show.

Where coffee grows

Coffee is concentrated in the south and south-west of the country. On the coffee authority's figures, as the USDA's attaché reports them, the regional state of Oromia grows about three fifths of it, and four states together, Oromia, South-West Ethiopia, Sidama and South Ethiopia, grow more than nine tenths. The report describes two belts, whose crops fared differently in 2025/26: a southern one that it lists as Sidama, Yirgacheffe, Guji and West Arsi, and a western and south-western one of Jimma, Limmu, Kaffa, Bench Maji and Sheka. Coffee has also been spreading into areas where it was not traditionally grown, such as Amhara in the north, though these remain a small part of the total. Some 5.9 million farmers grow coffee, by the authority's count, typically on less than half a hectare each and mixed with food crops, and smallholders account for nine tenths of production. Commercial farms cover 143,000 hectares, mostly in Oromia and the south-west.

CoffeeHQ states no altitude for any Ethiopian coffee region, because no document read gives one. What is sourced is the range of the wild plant: the Kew study, citing earlier work, gives the altitude of wild arabica across its whole native range, which reaches into South Sudan and northern Kenya, as 950 to 1,950 metres, with 1,200 metres the most usual lower limit. For climate, a field study in the Jimma zone gives an example of the pattern in the south-west: about 1,600 millimetres of rain a year, a long dry season from October to January in which coffee is picked and processed, and short rains from February to May in which it flowers.

Four ways of growing coffee, from forest to plantation

Researchers describe Ethiopian coffee as grown under four kinds of management. In forest coffee, the shrubs grow naturally under the canopy of highland forest with little or no human intervention. In semi-forest coffee, people thin the canopy and clear the undergrowth to give the coffee more light and less competition. Garden coffee is planted around the home, and plantation coffee is planted as a crop. The classification is older than the 2022 field study in the Jimma zone that is followed here; that study takes it from work published in 2008, which is also its source for forest coffee yielding only around 50 kilograms of green coffee a hectare a year, and cites other work, of 2009 and 2012, for the first two kinds together supplying 44 per cent of the country's coffee. A 2014 study of the export trade quotes one of the same sources for a fuller split, about 10 per cent forest, 35 semi-forest, 50 garden and 5 plantation, and warns that reliable recent statistics are lacking. None of these is a current count.

The study's own finding concerns disease. Across 68 sites, coffee leaf rust was significantly less severe in forest coffee than in semi-forest coffee, in both seasons measured. The authors attribute part of the difference to denser canopy and less human traffic, and suggest that the more uniform coffee planted in managed forest may also play a part. Their conclusion is a trade: more management gives more coffee per hectare, and proportionally more of it is lost to rust.

The government now wants a kind the classification has no place for. The USDA's attaché reports that 100,000 hectares have been allocated to private investors for large, mechanised coffee farms, reportedly in Oromia and the south-west, the first time large tracts have been set aside for the purpose, with Brazil cited as the model. In May 2026 planting had not begun. Some in the industry, the report notes, doubt that mechanised farming can keep the character of Ethiopian coffee.

What the names on the bag are names of

The words that identify Ethiopian coffee are of three different kinds, and they overlap. Some are regional states of the federation: Oromia, Sidama, South Ethiopia, South-West Ethiopia. These are what production statistics are reported by. Some are smaller places inside them. The USDA's report refers to Yirgacheffe as a district, and the Jimma zone, where much of the field research is done, is part of Oromia. And some are categories of the coffee trade.

The trade categories are the oldest. A study of every export transaction from 2006 to 2013 sorted Ethiopian coffee into Sidama, Jimma, Wollega, Yirgacheffe, Limu, Harar and other, and found that the first three made up about four fifths of exports by volume, while Harar and Yirgacheffe were small in volume and higher in price. In 2005 an institute was set up to secure legal ownership abroad of specialty coffee names, Sidamo, Harar and Yirgacheffe above all, as trademarks.

So a name such as Sidama can mean a regional state or a class of export coffee, and the two need not have the same edges. CoffeeHQ has read no document that defines the boundary of any Ethiopian coffee name, and for that reason gives none, and treats Yirgacheffe as part of this entry instead of as a region with a page of its own.

Species and varieties

Arabica only. What is grown falls into two broad kinds, and the trade's usual word for all of it, heirloom, names neither; a section below takes that word apart. The first kind is landraces: local populations selected and kept by farmers, often several together in one plot. A study of coffee forests in the south-west notes that farmers there have relied on a limited number of local landraces. The second is selections released by the national research system. One of them is 74110, which a processing study carried out at the Jimma research centre describes in passing as a well-adapted and previously released variety; CoffeeHQ has read no breeder's account of it. The national research institute, as the USDA's attaché reports it, says more than fifty improved varieties offering higher yields and better disease resistance have been distributed to growers. How much of the crop is one kind or the other, no document read says.

Where arabica comes from: what the genome says

Arabica is a hybrid. Each plant carries two sets of chromosomes, one from an ancestor of today's robusta and one from an ancestor of a second wild species, Coffea eugenioides. A 2024 study that sequenced the genomes of all three dates the event that joined them to between 610,000 and 350,000 years ago, long before anyone cultivated anything.

The same study sequenced 41 arabica plants, among them 17 collected wild from both sides of the Great Rift Valley. It found that the wild plants themselves have very little genetic diversity, which it attributes to two periods in which the population shrank, and that cultivated arabica has only slightly less. That is unlike most crops, where domestication discards most of the wild variety. An earlier study from Kew puts the contrast more strongly, saying the genetic diversity of wild arabica far exceeds that of cultivated varieties and of the collections kept by researchers, and that the wild plants vary in resistance to disease and tolerance of drought. The two studies measure diversity differently, and CoffeeHQ does not reconcile them.

On where the cultivated lineage came from, the genome study is careful. Typica and Bourbon, the two lines from which most of the world's arabica descends, are closely related to each other and share an origin in Yemen. The wild plants closest to them were found on the eastern side of the Rift Valley and around Gesha in the south-west. The authors say their data cannot identify the exact place. Their model has the lineage that became cultivated arabica splitting from the wild one about 30,000 years ago and exchanging genes with it for a long time afterwards, from which they infer that the two were separated by no great distance, perhaps the two sides of the Rift Valley; they also allow that the lineage may have reached as far as Yemen.

What heirloom means, and what it does not

Ethiopian coffee is very often sold with the variety given as heirloom. None of the documents read for this entry, which include the USDA's report on the sector, a study of the export trade and four scientific papers, uses the word at all, as the name of a variety or otherwise. That is an argument from absence, and CoffeeHQ has found no document that defines the word or records where it came from. The World Coffee Research catalogue points the same way: its index of arabica varieties has no entry called heirloom, and it files the Ethiopian material it does list, Geisha and Java among them, under the genetic description Ethiopian landrace. On that evidence it is best read as a word of the importing trade, and what it reports is that the seller does not know which plants the coffee came from.

That is usually an honest statement. A lot bought through a washing station is the cherry of many smallholders, each with a fraction of a hectare, and the plants may be local landraces, released selections such as 74110, or both. The research institute says more than fifty improved varieties have been distributed. Nothing on a bag distinguishes them.

Two things follow for a buyer. Heirloom does not mean old or wild: a lot so labelled may be a modern selection bred for yield and disease resistance. And it is not a flavour. Whatever distinguishes one Ethiopian coffee from another in the cup, the word carries no information about it. Whatever variety there is among Ethiopian coffee plants is a property of the country's forests and farms taken together, and the label on a bag does not report it.

Old trees, and a campaign to cut them back

Ethiopian yields are low: the USDA's attaché estimates about 0.9 tonnes of green coffee a hectare, on a harvested area of 790,000 hectares in 2025/26. One reason is age. Industry sources told the attaché that nearly 70 per cent of the country's coffee trees are old, some of them more than a hundred years.

The response has been stumping, cutting an old tree back to a stump so that it regrows. A national campaign began about four years before the report, and the coffee authority says stumped trees covered 15 per cent of the harvested area in 2025/26, with the highest rate in Oromia. Studies in Sidama and South Ethiopia, as the report describes them, found that yields can rise by up to three times within four years of stumping. The attaché attributes part of the recent growth in the harvest to it. The national research system is also multiplying improved seedlings by tissue culture; the hybrids it describes yield about 2.8 tonnes a hectare under good management, which is the institute's own figure.

Harvest

The main harvest runs from October to December on the USDA's account, and the marketing year is counted from October. Cherries are picked by hand; in Sidama in 2025/26 pickers were paid by the kilogram. Timing varies with place and year: the report describes the southern belt and the south-west as having had different fortunes in 2025/26, and a late start to the rains delaying the southern harvest by up to three weeks. The coffee does not leave when it is picked. Exports are weakest in December and January, while the crop is being processed and graded, and strongest from May to July.

How the coffee is processed

Both of the oldest methods are in wide use, and the natural method is the larger. For washed coffee, cherries are pulped at a washing station as soon as they are picked, fermented in tanks, washed clean and dried in the sun in their parchment. For natural coffee, called unwashed in Ethiopian trade documents, the whole cherry is dried on mats, raised beds or concrete and hulled afterwards. A 2024 study carried out at the Jimma research centre cites, without a year, a proportion of about 65 per cent dry-processed to 35 per cent washed, and a study of export records found washed coffee close to 30 per cent of exports throughout 2006 to 2013. Which is made depends on place and, as a section below explains, on the price of cherries.

Washed or natural: a choice that follows the price

Whether an Ethiopian coffee is washed or dried whole is partly a matter of place. In the export records for 2006 to 2013, coffee sold as Yirgacheffe was significantly more likely to be washed than Sidama; coffee from Jimma and Wollega was much less likely to be; and no washed coffee was exported from Harar at all. It is also a matter of who handles it: cooperatives, state farms and commercial farms were more likely to wash than private traders were.

And it moves with the market. Washing needs a station, water and money to buy cherries at harvest. The export study found washed coffee earning a clear premium abroad, yet its share of exports did not rise over the seven years of records. The USDA's 2026 report shows the mechanism working in reverse. After record prices, the price of fresh cherries at the start of the 2025/26 harvest tripled in some areas and quadrupled in others; several cooperatives in Yirgacheffe reported it peaking at nearly four times the year before. Washing stations were less able and less willing to buy at volume, and a number of farmers reportedly chose to dry their cherries at home and sell natural coffee themselves. The result the attaché reports is a tighter supply of washed coffee and a growing share of natural, in the south and the south-west alike.

Washed coffee also reaches the market sooner. The export study found it shipped earlier in the year than natural coffee, which takes longer to dry.

How it is graded and sold

Ethiopia sells its coffee almost entirely as green beans, and through a trade the USDA's attaché calls one of the most highly regulated sectors of the country's economy. Every exported lot is graded and certified, a minimum export price is set each week, and coffee graded for export cannot be sold at home. Coffee earns the country about a third of its export revenue, on the USDA's 2026 report. Its largest buyers in 2024/25 were Saudi Arabia and Germany, each taking about a sixth, followed by China, Belgium and the United States; China's purchases had more than tripled in a year, helped by tariff-free access. How lots are graded, and how the rules on who may sell have changed, are set out in two sections below.

How a lot is graded

Every lot for export is inspected twice, raw and in the cup. What follows is the system as a study published in 2014 recorded it, not the current rules. The raw inspection counted for 40 per cent of the grade and the cup for 60 per cent, the latter split equally between cleanliness, acidity, body and flavour. Before either, a lot had to pass two tests: moisture below 11.5 per cent, and 85 per cent of the beans retained on a screen of size 14.

The grades were then assigned by process. Washed coffee was graded 1 or 2 and natural coffee 3, 4 or 5, with an undergrade below each, so that on that system the grade number recorded the processing method as well as the quality. Raw washed coffee was judged on shape, colour and smell; raw natural coffee on a count of defects and smell.

The system has since been revised. The USDA's 2026 report describes export grades as 1 to 5 without tying them to process, and quotes the Sidama office of the coffee authority as classing 97 per cent of the coffee the region supplied to the central market in 2025/26 as grade 1 or 2. CoffeeHQ has not read the current grading rules themselves and does not describe them beyond that.

The exchange, and the routes round it

From December 2008 private traders in Ethiopia were required to sell their coffee through a new commodity exchange. It traded standard contracts against warehouse receipts, with set grades, lot sizes and terms of payment, and moved the first quality inspection out of Addis Ababa to nine units in the producing areas. Two kinds of seller were exempt from the start: farmer cooperatives, and producers who exported their own coffee. The state intervened forcefully in the same years. In 2009 it revoked the licences of six large exporters accused of hoarding and sold their stocks, and from 2011 it limited how much coffee an exporter could hold without a contract to ship it.

The exemptions have since widened into the main road. A proclamation of 2017 allowed producers to take part directly in processing and export, and the USDA's 2026 report describes the result: exporters, cooperatives and qualifying producers buy from farmers or washing stations and export without passing through the exchange floor, which keeps a lot traceable to where it was grown. A farmer with two hectares or more may now export directly. The report describes this as reduced reliance on the exchange, not its abolition, and directives of 2025 raised the capital an exporter must hold and opened raw coffee export to foreign investors.

One instrument has not changed hands. The coffee authority sets a minimum export price, updated weekly by grade, process and origin, below which a contract cannot be registered. There is no ceiling.

Coffees from this origin

How coffee is drunk in Ethiopia

Ethiopia keeps a large share of its coffee for itself. The USDA's attaché estimated domestic consumption at 4.5 million bags in 2025/26, close to two fifths of the harvest, and forecast 5.0 million for 2026/27; it puts consumption per person at about two kilograms a year. A 2014 study by the International Food Policy Research Institute, citing the national statistics agency, put the share higher still, at more than half of production. Both documents explain it the same way: coffee is part of daily social life, anchored in the buna ceremony, in which coffee is roasted, ground and brewed in a clay pot called a jebena. The attaché describes a large informal trade of street vendors serving jebena buna alongside a fast-growing café trade in Addis Ababa and other cities. The two markets are also separated by law. Coffee graded for export may not be sold at home, and what fails the export standard is what the domestic market receives. The ceremony has its own CoffeeHQ entry, linked below.

The two shares are from different years and bodies and are not a trend: the USDA's is its own estimate for 2025/26, and the 2014 study's is the statistics agency's figure for about 2013.

Preparation traditions

How coffee is traditionally prepared and served in this country, in CoffeeHQ's international preparation atlas.

• Ethiopian coffee ceremony (jebena buna)
• Eritrean coffee ceremony (bun)

History

Ethiopia's place in coffee history rests more securely on botany than on documents. William Ukers, in his trade history of 1922, accepted without reserve that coffee was first made known from the mountains of north-eastern Africa, and he did so in the absence of any conclusive evidence to the contrary, not on the strength of a record. The story usually told, of a goatherd called Kaldi, is a legend, treated below. By Ukers's time what was then called Abyssinia exported only a little: a cultivated coffee from around Harar, sent by mule and railway to Djibouti and on to Aden, and a coffee gathered from wild trees in Sidamo, Kaffa and Guma. A century later it is the largest producer in Africa, and its harvest has more than doubled in the past twenty years.

The goatherd, and what the record actually holds

The story of Kaldi, the goatherd who noticed his goats dancing after eating coffee cherries and took the discovery to a monastery, is the most repeated account of coffee's origin. Ukers reproduces it as a legend, giving the goatherd's name in what he calls the French version of it. He does not say where the story was first printed. What he does record is that Antoine Faustus Nairon, a Maronite professor in Rome whose Latin treatise of 1671 he calls the first devoted to coffee alone, accepted the tale of the Abyssinian goatherd. It was doubted early. Ukers reports the scholar Galland, whose analysis of an Arabic manuscript he treats as the most trustworthy account of coffee's origin, criticising Nairon for accepting the tale of the Abyssinian goatherd; Galland called it unworthy of belief as a fact of history, while allowing that there might be some truth in it.

What can be said with evidence is narrower and more interesting. The plant is native to the Ethiopian highlands; the genome evidence places the origin of cultivated arabica there or close by, without fixing the place; and the two lines that left for the rest of the world did so through Yemen. When, by whom and in what form coffee was first used in Ethiopia, no document CoffeeHQ has read establishes.

Pressures on the sector

• For the wild plant, the outlook modelled so far is poor. A 2012 study led from the Royal Botanic Gardens, Kew, modelled where wild arabica could grow, using one climate model and three emissions scenarios. Its most favourable result was a fall of about 65 per cent by 2080 in the number of known wild localities that remain climatically suitable, and its least favourable an almost complete loss; measured as area, the range was a 38 to 90 per cent reduction. It is a model of climate suitability, not a count of trees, and the authors say so.
• For cultivated coffee, a more recent projection points the other way. The USDA's attaché reports that Ethiopia's coffee industry is citing a study published by a bank in March 2026, which projects the country's suitable arabica area growing by 2050 while Brazil's, Colombia's and Honduras's shrink. CoffeeHQ has not read that study and reports it second-hand.
• In the meantime the risk is the weather of the year. The attaché names irregular rain and drought as standing concerns, and reports that in 2025/26 a late start to the rains disrupted flowering in the south and delayed the harvest by up to three weeks, most of all at high altitude.
• Old trees and low yields. Nearly 70 per cent of trees are reported to be old, and the national average is under a tonne a hectare.
• The cost of cherries. In 2025/26 record prices left washing stations unable to buy and exporters holding stock they could not sell at a profit.
• Proving where coffee was grown. The European Union's deforestation regulation requires plots to be mapped, across millions of farms of less than half a hectare; hundreds of thousands had been registered by 2026.
• Organic certification. The European Union's stricter rules, binding from 2025, raised inspection rates and capped group sizes, and cooperative unions have cut the number of certified cooperatives.
• The wild plant. The populations that hold arabica's remaining diversity are modelled to lose most of their climatically suitable ground this century.

The two projections are not in conflict so much as about different things: one is of wild populations in forest, the other of land that could be farmed. Neither is an observation.

Institutions worth knowing

• Ethiopian Coffee and Tea Authority (ECTA) — the regulator, which registers the trade's transactions, oversees the grading and certification of export coffee and sets a weekly minimum export price
• Ethiopia Commodity Exchange (ECX) — the exchange through which private traders were required to sell coffee from December 2008, and which is no longer the only channel
• Jimma Agricultural Research Center — an agricultural research centre in the south-west where the processing study cited here was carried out, on the released variety 74110
• Ethiopian Institute of Agricultural Research (EIAR) — the national research institute, which the USDA's attaché reports as saying more than fifty improved coffee varieties have been distributed

Explore Ethiopia's coffee from here

A reading order, not a list of everything related: each step picks up something this page raised and takes it further.

1. Step 1Arabica, the species that began hereWhat kind of plant Coffea arabica is, and how it differs from robusta in the field and the cup.
2. Step 2Ethiopian heirloomThe short entry for the trade word this page explains at length.
3. Step 3GeshaThe variety named after the south-western area where the genome study found wild plants among the closest to cultivated arabica.
4. Step 4YemenWhere the two lines that became most of the world's arabica were first cultivated.
5. Step 5Washed against naturalWhat each of Ethiopia's two methods does to the same cherry, side by side.
6. Step 6Natural processingThe method most Ethiopian coffee goes through, on its own page.
7. Step 7The Ethiopian coffee ceremonyHow the coffee Ethiopia keeps for itself is roasted, ground, brewed and shared.
8. Step 8Early coffee historyWhat is documented about coffee before it reached Europe, and what is not.
9. Step 9KenyaThe neighbour to the south, whose northern mountains hold an outlying population of wild arabica.
10. Step 10Coffee Expert: species and varietiesThe lesson on what a species, a variety, a landrace and a selection each are.
11. Step 11Tasting LabTaste a washed and a natural Ethiopian coffee together and write down what you find, unprompted.