Dates, figs, a heap of sun-dried apricot. Pour a cup of Ja'adi and the fruit lands first, then a proper wallop of cardamom and dark chocolate arrives underneath it. Dense, syrupy, faintly wild, and completely its own thing.
That makes a certain kind of sense, because Ja'adi comes from Yemen, the country where coffee stopped being a plant and started being a drink.
The twist is that Ja'adi probably isn't a variety at all, and the story of how that came to light is a good one.
A name that describes a tree
Start with the word. Ja'd in Arabic means curly, or rough. Ask a farmer in the mountains west of Sana'a why their trees are called Ja'adi and you'll get one of two answers. Some point at the leaves, which have a wavy, crimped edge that catches the light differently to their neighbours. Others point at the cherries, which grow in tight, knotted clusters pressed hard against the branch. Lovely to look at, an absolute pain to pick.
Both answers are about appearance, and ancestry doesn't come into it. It's the way a smallholder anywhere might call the tree at the bottom of the orchard "the knobbly one", and it's how many Yemeni coffee names work. Tufahi comes from tuffah, apple, because the cherries are apple-shaped. Dawairi means rounded, probably a nod to bushy growth. Other names are borrowed from the places the seed came from.
For decades the specialty trade treated these names as varieties, printing them on bags right where you'd otherwise expect Bourbon or Caturra. A reasonable assumption, and a wrong one.
How coffee got to Yemen
Arabica isn't native to Yemen. Its natural habitat is the southwestern highlands of Ethiopia and the Boma Plateau of South Sudan. Yemen is where the plant was domesticated, and that distinction matters a great deal.
Coffee arrived across the Red Sea somewhere around the 1400s. Sufi communities in western Yemen were brewing it to stay alert through all-night devotions, and the drink spread outward from there through the port of Al Mokha, a name that stuck to the coffee long after the port itself faded. Yemen held the global trade for roughly two centuries and guarded it, with beans often boiled or parched before export so nobody could plant them elsewhere. Then Dutch traders got seed out, planted it in Java, and much of the coffee world was built from a genetic handful.
Typica, Bourbon and the varieties that followed them, Caturra, Catuaí, Mundo Novo and Pacas among them, all trace back to that handful. So the trees still standing on Yemeni terraces today belong to the population that much of the coffee world grew out of. Same soil. Same families, farming coffee for four or five centuries without a break.
What the DNA actually said
A team funded by the Qima and Lavazza foundations went looking for an answer to a question nobody had properly asked. Do Yemeni coffee names mean anything genetically?
They visited 148 farms and asked farmers to name what they were growing. Eighteen different names came back. Udaini dominated at 34% of citations, followed by Dawairi at 16% and Tufahi at 14%, with Buna and Ja'adi trailing well behind.
Ja'adi turned out to be very local. It was specific to one governorate, Sana'a, and within that it clustered in Haraz. Nobody in Ibb, Dhamar, Raymah or Al Mahwit called their trees Ja'adi at all.
Then came the genotyping. The team sampled 88 trees, including 18 identified as Ja'adi by the farmer standing next to them, and ran them through microsatellite markers. The result, published in Agronomy in 2022, was blunt. There was no significant association between the name a farmer gave a tree and the genetic group that tree belonged to. Udaini samples turned up across all three of Yemen's known genetic groups. So did Tufahi. The names simply weren't tracking the genes.
Ja'adi came out better than most, mind you. Of the 18 samples, 15 belonged to the New-Yemen group, two to Typica/Bourbon and one to SL34. That's about 83% in a single group. A decent bet, well short of a guarantee.
The finding that reframes everything is this one. Where a tree was growing predicted its genetics far better than what it was called. Region beat name, comfortably.
Farmers weren't getting it wrong. Two naming systems were built for different jobs. Yemeni farmers were describing trees they could see, centuries before anyone could read a genome. Trouble only starts when the specialty market treats those descriptions as pedigree, and when farmers buy seedlings by name expecting particular traits.
Yemenia: the third mother
The Ja'adi result reads as good news, and that comes down to a discovery made two years before the study was published.
In 2020, after a genetic survey covering more than 25,000 square kilometres, Qima and Dr Christophe Montagnon announced an entirely new mother population of arabica, growing only in Yemen. They named it Yemenia, from the Arabic for "the Yemeni mother". The 2022 study calls the same group New-Yemen, so if you go looking for the paper, that's the word to search for.
Until then, the cultivated arabica world was explained by two mother populations: Typica/Bourbon, identified through the seed movements of the 1700s, and the SLs, catalogued in the 1900s. Yemenia is a third, and it exists nowhere else on earth.
The obvious question was whether it tasted like anything. Qima partnered with the Alliance for Coffee Excellence and put the coffees in front of a blind jury of 35 judges from 24 companies across 14 countries. Many lots scored 90 and above. One judge called it "explosive", and another found the berry notes of an SL28, the sweetness of a yellow Bourbon and the aromatic complexity of a Geisha.
So when a farmer in Haraz hands you a bag marked Ja'adi, there's a strong chance you're holding Yemenia. Not certain. Strong.
The Ja’adi family tree
Ja’adi
The name at the centre of this article, from the Arabic ja’d, meaning curly or rough. Some farmers say it describes the wavy leaves, others the tightly bunched cherry clusters that make picking such hard work. Ja’adi is one of the most geographically specific Yemeni names, cited only in Sana’a Governorate and concentrated in Haraz. Genotyping in 2022 found 15 of 18 sampled Ja’adi trees in the Yemenia group, with two Typica-Bourbon and one SL34. A strong indicator, then, but not a genetic identity.
What a Ja'adi tree actually looks like
Ja'adi covers more than one genotype, so there's no definitive Ja'adi tree. What there is is a set of traits farmers in Haraz recognise on sight, and growers there draw further distinctions of their own, including an eastern Ja'adi found in Sharqi Haraaz.
The leaves have that wavy, crimped margin the name refers to, noticeably different from the flat, elongated leaves farmers elsewhere associate with Udaini. The cherries grow in dense clusters gripped tight to the branch, and they're small. So are the beans. Hard, compact, often slightly irregular in shape. Yemeni beans have always been undersized by the standards of the screen sizes the rest of the industry sorts by, which is one reason Yemen has historically struggled to fit into conventional grading systems.
Then there's the ground it grows on. Haraz sits in the mountains west of Sana'a, on stone terraces built by hand generations ago to stop the soil sliding down slopes that really shouldn't be farmable at all. Elevations run from roughly 1,700 to 2,600 metres. Manakha, the commercial hub of the Haraz trade, sits at around 2,200. The soil is volcanic and mineral rich with sharp drainage. Days are hot, nights are properly cold, and that diurnal swing slows cherry development to a crawl. That's where a lot of the density and sugar concentration comes from.
The really striking figure is rainfall. Yemen's coffee regions get under 400mm a year, a fraction of what arabica usually gets elsewhere. Haraz has an advantage in its traditional water system, built around stone dams, covered cisterns and spate channels that capture flood water, but the margins are thin everywhere. There's no chemical fertiliser and no fungicide. Farmers use animal manure and finely ground stone scattered around the base of the trees to feed the soil and hold moisture. It's organic because that's how it has always been done, in a place where inputs are expensive and water is precious, and marketing has nothing to do with it.
The cup
Almost all Yemeni coffee is natural processed, and again, that's down to necessity rather than fashion. You can't wash coffee where water is the scarcest thing there is. Whole cherries dry in the husk with the fruit intact, on rooftops and increasingly on raised beds, sun-baked slowly until the fruit turns to leather around the seed.
The result is a cup with a particular kind of gravity. Dried fruit first and in volume: date, fig, raisin, plum, sun-dried apricot. Then a darker fruit layer underneath, blueberry and black cherry, more jammy than bright. Then the thing that really marks Yemen out, which is spice. Cardamom, clove, nutmeg, a touch of cedar or tobacco, sometimes something almost incense-like. Underneath everything sits dark chocolate and molasses. The body is thick and syrupy, the acidity gentle and low-slung rather than sparkling, and the finish goes on for an age.
The cleaner lots push further. Where cherry selection and drying are tight, red fruit starts coming through on top of the dried: raspberry, redcurrant, sometimes a lick of blackcurrant, with the sweetness turning towards honey on the finish rather than settling into molasses. Same fundamental character, better resolution.
If you know natural Ethiopians, the contrast helps. Ethiopian naturals tend to run bright, floral and berry-forward, with a lifted, almost perfumed quality. Yemen goes the other way. Deeper, denser, spicier, more savoury-sweet, and it drinks more like a fortified wine than a fruit juice. Same species, same ancestral stock, very different result. That's what centuries of drought and altitude do to a plant.
Worth being straight about quality, too. Yemeni coffee has historically had a mixed reputation, and not unfairly. Rooftop drying with limited cherry sorting, long storage and blending along the way produced plenty of lots that were rustic at best and unpleasantly fermented at worst. What's changed in the last decade is the arrival of raised beds, which let air circulate under the cherry, along with tighter picking and sorting and cash paid at the farm gate for fresh cherry rather than dried. The best modern Yemeni lots are very clean. That's a processing story, not a genetics one.
For brewing, filter shows Ja'adi at its best. The clarity gives the spice and dried fruit room to separate out rather than collapsing into general sweetness. It works as espresso if you like a heavy, fruit-forward shot, though it can turn boozy if you push the ratio. Whatever you do, don't roast it dark. The spice is the point, and it's the first thing to burn off.
So if it's that good, why isn't it everywhere?
A few reasons, and the cup isn't one of them.
The first is arithmetic. Yemen is a tiny slice of global coffee production. Something in the region of 120,000 smallholders farm about 35,000 hectares, producing near 21,000 tonnes of cherry in a typical year. Yields sit at about 0.6 tonnes per hectare, a fraction of what a Brazilian or Colombian farm would consider viable. Plots are tiny. Around 40% of what's grown is drunk at home, and only a sliver of the rest reaches Europe or North America.
The second is that Yemen has been at war since 2015. Farms have been damaged, roads cut, farmers displaced, and exporting anything at all has become an exercise in persistence. Coffee also competes for land and, critically, for water with qat, which pays reliably and quickly in a way coffee historically hasn't. That's worth qualifying, though: in parts of Haraz the trend has gone the other way, with villages pulling out qat and replanting coffee terraces. It's one of the few hopeful stories in Yemeni agriculture.
The third is traceability, and it's structural. Exporters mostly don't buy from farms. Coffee is bought in the pod as whole dried cherry through a long chain of middlemen, stored, and blended along the way, often mixing farms, villages, varieties and sometimes crop years before it reaches an exporter at all. Traditional market names like Matari, Sanani and Harazi describe broad areas rather than specific places, and even those get used loosely. That's why a Yemeni coffee traceable to a single farm is still rare, and why it usually means someone has built a route that bypasses the traditional trader network, buying fresh ripe cherry direct and gathering it through their own collection centres.
The fourth is the one this whole article circles back to. There's no functioning seed sector. You can't walk into a nursery and buy verified Ja'adi the way you can buy verified Caturra, because nobody can verify it. There's no genetic definition to verify against.
That matters far more for Yemeni farmers than for the rest of us. When a farmer replants and asks for Ja'adi seedlings expecting particular traits, they may get Yemenia, or Typica-Bourbon, or SL34, with meaningfully different behaviour in the field. Get that wrong at scale and you've committed years of a smallholding to trees that don't suit the site. That's why the recommendation coming out of the 2022 study is to build genetically characterised collections in Yemen itself.
Origin expressions
Since region turns out to predict genetics better than name does, the map is the thing to know.
Haraz, in the mountains west of Sana'a, is where Ja'adi actually lives and is widely reckoned to be Yemen's finest growing area. Coffees here lean towards deep berry sweetness, wine-like complexity and a thick syrupy body, with dark stone fruit and dried cherry running through them. Bani Matar, next door and higher still, is the source of the coffee traditionally sold as Mocha Mattari, and tends towards dark fruit, tamarind and a wilder, earthier spice. Sanani is the broader catch-all for the highlands around the capital, historically the most commercially traded and generally a touch softer and more chocolate-led.
Ja'adi as a name belongs to Haraz. If you see it on a bag from anywhere else in Yemen, it's worth asking a question or two.
Why the name still matters
Ja'adi is a small, hard-won, deeply strange and thoroughly delicious coffee, with an asterisk attached.
It's a description of curly leaves and stubborn cherry clusters, handed down through generations of farmers on terraces in Haraz, rather than a variety in the way Caturra or Geisha are varieties. It also happens to correlate fairly strongly with the newest and least explored mother population in arabica. That's no flaw in the name. Farmers were classifying these plants long before anyone had a microsatellite marker, and their system was built to answer different questions.
Drink it for the spice and the sun-dried fruit and that impossibly long finish. And drink it knowing the trees in your cup are standing where the whole thing started, in soil that shouldn't really support coffee at all, tended by families who have been doing this since before Bourbon reached Brazil.
Further reading
Agronomy: Vernacular Names and Genetics of Cultivated Coffee in Yemen
The open-access 2022 study behind most of this article. Covers 148 farms, 88 genotyped trees and the full breakdown of which names map to which genetic groups. The table of name origins is worth the visit on its own.
Genetic Resources and Crop Evolution: Unveiling a Unique Genetic Diversity of Cultivated Coffea arabica in Yemen
The paper that described the New-Yemen genetic group, later named Yemenia, and confirmed Yemen's role as the domestication centre of arabica.
Qima Coffee: Yemenia
The producer-side account of the discovery, including the Alliance for Coffee Excellence blind cupping results and the first international auction of Yemenia lots.
World Coffee Research: Arabica Variety Catalog
There's no Ja'adi entry, which is precisely the point. Useful for comparing how thoroughly documented varieties like Bourbon, Caturra and Geisha are described against what we know about Yemeni landraces.
FAO: Coffee Value Chain in Yemen
Production figures, land area, export values and a clear-eyed look at the challenges facing Yemeni coffee farming.
Curious how Yemen's seeds became the rest of the coffee world? Read our pieces on Typica, SL-34, and Geisha.