You cannot buy, brew or taste affinis coffee. Coffea affinis is a wild species of the Upper Guinean forests of West Africa that effectively vanished from the botanical record for generations before being relocated in the wild in recent years, and it remains a plant nobody farms, grades or brews. What it has instead of a cup is one of the most useful ideas in the whole genus: it shows exactly what people mean, and do not mean, when they say a coffee species was "lost".
One correction first, because it governs everything that follows. Coffea affinis is a species. It is not a variety and not a cultivar. Bourbon, Typica, SL28 and Geisha are varieties — genetic variants inside the single species Coffea arabica, the way different apple cultivars are all one apple. Coffea affinis is not inside arabica at all. It sits on its own branch of the genus, and every instinct carried over from variety talk has to be put down at the door.
What affinis coffee actually refers to
It refers to a forest shrub, not a beverage and not a trade category. The published record places Coffea affinis in the Upper Guinean forest belt: collection evidence establishes it as an indigenous wild plant of Guinea and Côte d'Ivoire, and in recent years it was located growing wild in Sierra Leone for the first time, in hill forest reported to be not far from populations of its better-known relative. Its recorded habitat is humid evergreen forest and gallery forest — the ribbon of woodland following watercourses — rather than the high, cool ground most drinkers associate with fine coffee.
Two physical details recur consistently. The ripe fruit is reported as black, or black-purple, rather than the red that arabica and robusta ripen to. And the feature separating it in the field from its close relative is vegetative rather than reproductive: broader, larger leaves, along with differences in stems and stipules, against the narrow foliage of the other species. That is a thin set of characters to hang an identification on, and a large part of why the two have been tangled together for well over a century.
Beyond that, the documentation is sparse: a handful of old collections, a few colonial-era agricultural notes, and a recent burst of field and laboratory work. There is no long run of observation, no accumulated grower experience and no body of measurement of the kind the farmed species have.
"Lost" is a statement about collectors, not about plants
This is the part worth getting right, because the same phrasing turns up every time a coffee species reaches the news. When coverage says a species was lost for a century and then rediscovered, what is nearly always meant is narrower and stranger than it sounds: for that stretch of time, no botanist filed a specimen collected from a wild population. The plant itself was very probably standing there the whole while. It was almost certainly known to the people living around that forest, who would have had their own name for it and their own sense of where it grew. What lapsed was not the species. What lapsed was the paperwork.
The mechanism is simple once you see it. A species officially exists, for scientific purposes, as pressed and dried material held in a collection, with a label recording who gathered it, where and when. That sheet is the unit of evidence. So a map of where a wild coffee grows is not really a map of the plant — it is a map of where botanists have walked, and how recently. Forests that are hard to reach, or that have not been anyone's project for decades, generate no sheets, and on the ledger the plant disappears.
Read that way, a rediscovery is a story about search effort rather than resurrection. Someone went back to a forest belt that had gone unsurveyed for a long stretch and found the plant where the old labels suggested it might still be. The reported gap between the last confirmed wild collection and the recent ones runs to the better part of a lifetime, and the question that gap answers is not "where did the species go" but "who was looking, and where".
It also explains why the coffee genus keeps producing both brand-new species and dramatic returns from the dead: a great deal of the genus lives in tropical African and Malagasy forest that is thinly collected, so the register is permanently out of date in both directions. The accurate translation of the next such headline is that a collection record now exists where one did not before. That is genuinely valuable — it is how conservation and breeding work get anchored — but it is a fact about human attention, not about a plant coming back.
Where it sits in the genus, and why it gets confused
The genus Coffea contains well over a hundred described species, and the count keeps moving as forests are surveyed and old names re-examined; our guide to how many coffee species there are covers that family picture. Two species carry essentially the entire world crop — arabica and canephora, the species behind robusta — with liberica a distant third. Coffea affinis has never been part of that picture.
Its nearest neighbour in the literature is Coffea stenophylla, the other forgotten West African species, which is the one with the famous return-and-tasting story and the heat-tolerance interest; that account belongs to our guide to stenophylla coffee. What matters here is the relationship. The two grow in the same forest belt, were reported growing wild near one another in the recent field work, and share the unusual dark fruit. More tellingly, in the coffee gardens of the region around the turn of the twentieth century the two are reported to have been grown side by side and moved under a single shared regional name — Rio Nunez coffee — rather than as two distinct species. All of that invites confusion, and confusion is a large part of what the record contains.
Here the sources genuinely conflict, and both positions deserve stating. Earlier authors proposed that Coffea affinis was not an independent species at all but a hybrid between liberica and stenophylla — a reading that recent work reports as ruled out on several grounds, among them that the plant sets fertile seed, that its fruit colour is consistent rather than intermediate, and that its molecular placement does not behave the way a hybrid's would. A separate and quite different argument holds that it might reasonably be folded into C. stenophylla at a rank below species rather than standing on its own; an old subordinate name of exactly that kind sits in the synonymy to this day, and at least one earlier author treated the two as the same plant. The current treatment keeps Coffea affinis as a species in its own right, but a reader should know that its rank has been argued over, that its range has been reported differently at different times, and that some historical material filed under one of these names may belong to the other.
For the country context around the most recent wild records, see our guide to Sierra Leone coffee.
At a glance
A qualitative comparison only — there are no reliable performance figures for this species, and none are invented here.
| Trait | C. affinis | C. stenophylla | C. arabica | C. canephora (robusta) |
|---|---|---|---|---|
| Rank | Species — not a variety of anything | Species | Species | Species |
| Range as reported | Upper Guinean forest — Guinea and Côte d'Ivoire, plus a recent wild record in Sierra Leone | Upper Guinean forest of West Africa | Highland forest of Ethiopia and adjacent areas | Central and West African lowland forest |
| Habitat | Humid evergreen and gallery forest in hill country | Lowland to mid-slope forest | Cool, higher-elevation forest | Warm lowland forest |
| Ripe fruit | Reported black to black-purple | Black to purple-black | Red, or yellow in some selections | Red |
| Leaves | Broader and larger than its close relative's | Narrow, lance-shaped | Broad, glossy, leathery | Large and broad |
| Domestication | None at all | Historically farmed, then largely abandoned | Centuries of selection | Well over a century of selection |
| Modern sensory record | None that is reliable or species-specific | A recent published evaluation exists | Extensive | Extensive |
| Commercial status | Not a crop; unavailable in any form | Reported as not under commercial cultivation | Most of world production | Nearly all the rest |
| Role today | A wild relative; a conservation and breeding question | A research and revival subject | The reference coffee | The volume coffee |
How the plant and its seed differ from a crop coffee
The differences that matter are not the ones an identification key lists. They are the differences between a wild forest plant and a plant shaped, over generations, into something a farm can run on.
A crop coffee has been selected for traits that have nothing to do with survival in a forest: cherries that ripen together rather than over months, so a picking round is worth making; branches that hold fruit where a picker can reach it; seed of a consistent size and shape, because pulpers, hullers, screens and roasters are all built around consistent seed; enough fruit per tree to justify the ground it occupies; and tolerance of being pruned, crowded and fed. Wild plants have none of that. They are selected for shade tolerance, for getting seed dispersed by animals, and for outlasting bad years — a completely different specification.
On the specifics of the seed, the honest position is that the public record does not support a confident description of size, uniformity or chemistry. What can be said is what has not been done: no one has run the generations of selection that would tell you which individuals in a wild population fruit heavily, ripen evenly or hold a bean size worth grading. Those questions have never been asked of this species, so any figure offered for them is guesswork.
One structural point is worth adding, because it explains why wild coffees so rarely convert into crops. The known wild material is small — a few forest patches, and populations described as tiny where they have been counted at all. A breeding programme needs variation to choose from, and a handful of plants supplies very little. Working from that starting point is a decades-long undertaking before anyone even asks what the seed does in a roaster.
What affinis coffee tastes like — the honest answer
Nothing reliable is known, and that is the plain state of it.
There is no modern published sensory evaluation of this species: no cupping score, no descriptor set, no panel result, no body or acidity note that can be attributed to it. Any tasting note offered for it is extrapolation or invention.
Two temptations are worth naming. The first is old testimony: century-old agricultural writing did speak favourably of the quality of coffee associated with this plant, and that remark still circulates. But material from that region and era moved under shared trade names covering more than one species, so a compliment of that kind cannot be pinned to Coffea affinis — and praise recorded generations ago by someone promoting a crop is weak evidence at the best of times.
The second is transference. Because its close relative has a well-publicised modern tasting result, it is easy to assume the two cups resemble each other. They are different species, no comparative evaluation of the two exists, and flavour is not inherited across a species boundary by association. Borrowing one species' tasting notes for another is the commonest error in coverage of this plant, and it is worth resisting even when the two names appear in the same sentence.
Why nobody grows it
Being interesting to science and being viable as a crop are unrelated conditions, and the gap between them is measured in decades of slow work.
To turn a wild coffee into a farmable one you need, roughly in order: enough planting material of known origin; a propagation method that reliably produces more of it; some idea of how the plant responds to spacing, shade, pruning and nutrition, which nobody alive has learned for this species; several generations of selection to lift yield and even out ripening; a processing protocol suited to its particular fruit; grading standards, so that a lot can be described at all; and buyers willing to handle an unfamiliar species. Every step takes years, because tree crops answer questions on the timescale of a tree.
None of that has been done here. The species has had no domestication at all — not a shortened programme, not an abandoned one, none. That is why it is accurate to call it a wild plant rather than a lost crop, and why anyone offering a timeline for when it might reach a cup is making it up. Its close relative, by contrast, at least has a farming past to draw on, which is one more reason the two should not be discussed as interchangeable.
Why it matters anyway
Two reasons, and neither depends on it ever being planted.
The first is breeding. The farmed coffee gene pool is famously narrow, and the traits a warming coffee belt will need — tolerance of heat, of long dry spells, of pests and diseases moving into new ground — have to come from somewhere. Wild relatives are that somewhere. A species does not need to be farmable itself to be valuable in that role; it needs to exist, to be identifiable and to be accessible for study, and the recent work on this plant is best understood as securing exactly that. Material held in living collections outside the forest is part of the same insurance, because a species known only from a wild patch has no backup.
The second is that the distinction between administrative loss and real loss cuts both ways. If a species is missing from the record only because nobody was looking, it can be found again — that is the good news in this story. But a plant known from a small number of forest patches can also be lost for real, and forest clearance does not care about the ledger. The extinction-risk picture across the genus is covered in our guide to endangered coffee species; the standing of this particular plant is best described as under-documented rather than confidently placed, which is itself a finding worth taking seriously rather than a gap to shrug at.
A quieter point sits underneath both. The people living around these forests were not waiting for a rediscovery: local knowledge of which plants grow where routinely predates the formal record by generations, and a species can be perfectly well known to a village and entirely absent from a herbarium at the same time.
The bottom line
What is reasonably documented about Coffea affinis: it is a distinct wild species of the Upper Guinean forest belt, recorded from Guinea and Côte d'Ivoire and found in the wild in Sierra Leone in recent years; its fruit ripens dark rather than red; it is separated from its close relative mainly by broader leaves; and its rank and relationships have been argued over for well over a century.
What is not established: anything about how it performs as a crop, and anything at all about how it tastes. The shortage of information is the true state of knowledge, not a space to be filled with a plausible guess.
The idea worth keeping is the one about the word "lost". Species are not misplaced; records are. Every time the genus produces a headline about a coffee returning from oblivion, what has really changed is that someone walked into a forest with the time and the reason to look — a fragile way to track a plant's existence, and a good argument for surveying the rest of the genus before the forests answer the question for us.