Gesha: Not one variety.

Section 01F1 ↑

1. At a glance

This chapter only lists the readings, and each number gets a chapter of its own later on. The name has three spellings, Geisha, Gesha and Gori Gesha, and each spelling has a different history behind it, which chapter 02 goes through. World Coffee Research (WCR) is a non-profit that studies coffee varieties, and its variety catalog is a standard reference in the trade. The catalog has just one Gesha entry, called “Geisha (Panama)”.

Gesha's number, T2722, comes from an entry in a garden register dated 28 July 1953. The garden that received it was the coffee introduction garden that the Inter-American Institute for Cooperation on Agriculture (IICA) ran at Turrialba in Costa Rica, and CATIE looks after that collection today. CATIE itself was not founded until 1973, so the common line that it “went into CATIE in 1953” is not accurate. The literature writes the number as T2722, T.2722, T.02722 and T-2722, and all of them refer to the same material.

There are two accounts of the year it became famous. The Esmeralda farm's own account and the 2004 results table of the Specialty Coffee Association of Panama (SCAP) agree with each other, and both point to 2004. As of 26 September 2026 the WCR catalog still says 2005, and a 2024 genome paper repeats that 2005 after it. Our timeline uses 2004 and records 2005 alongside it.

Two prices

Best of Panama (BOP) is the competition and auction that SCAP holds every year. In 2025 its lot GW-01, 20 kg of washed Gesha from Hacienda La Esmeralda, sold for $30,204 a kilo.

We checked the auction platforms, the Cup of Excellence results pages for 11 countries, and reporting in Chinese, English and Japanese. As of 26 September 2026 we had found no higher sale price on public record. Privately negotiated sales and platforms inside China were outside the scope of our search.

The top price in 2026 was $18,004 a kilo, which did not break the record. But the median price of that year's 40 Gesha lots was $3,019 a kilo, against only $1,166.50 in 2025.

The median is the price in the very middle when all the lots are lined up by price; when the number of lots is even, it is the average of the middle two. The record is one lot; the market is all the others. Chapters 08 and 09 take this further.

Section 02F2 ↑

2. One name, three things

The catalog says it first

In the Geisha entry of its variety catalog, WCR writes that “there are multiple genetically distinct plant types that have been referred to as Geisha”. The sentence has not changed by a single word since 2023.

Here a variety catalog says, unprompted, that one of the names it lists does not point to one consistent thing. This dossier starts from there.

Three things

The first is the line that can be traced to T2722, and the Panama Gesha that WCR describes comes from it. The second is the wild population in the forests around Gori Gesha in south-western Ethiopia, which was never uniform in the first place. At the source, the Gesha Village farm grows a line specifically to keep this forest's genetic diversity, as chapter 05 explains.

The third is material sold elsewhere under the Gesha name, some of it with a well-documented history and some with very little. A study published in July 2026 found one case in Ecuador. A farm had paid $500 a kilo for “Gesha” seed, and when it was tested with 11 SSR markers it turned out to be a different Ethiopian landrace, much like T.02722 but not the same.

SSR markers are short sequences repeated along stretches of DNA, and they are commonly used to check whether two samples share a source. The paper's authors call this material “Evangelina” for now, and note that the name is used only for tracing within the farm.

Where the 39% comes from

In 2020 Pruvot-Woehl and colleagues tested 2,533 samples with 8 SSR markers, and 88 of them were held as Gesha in the WCR database. Of those, 39% matched CATIE's reference plant T.2722 exactly, 24% were very close and belonged to the same Gesha genetic group, and 37% had a genetic background unrelated to Gesha. WCR's 2020 press release cites the same 39%.

These 88 samples were ones sent in to the WCR database, not drawn in proportion to the market, so the 39% cannot be read as “61% of the Gesha on sale is fake”. What it does show is something else. Until it has been tested, material called Gesha cannot be assumed to be T2722.

The spellings

WCR's explanation is that the earliest germplasm records spelled this material Geisha, and research institutions have kept that spelling ever since. The mountain near where it was collected is usually spelled Gesha in English, and the two spellings are interchangeable.

Gori Gesha is a forest in the West Omo Zone, a different place from the Gesha district in the Kafa Zone. The Gesha Village home page says West Omo, while its page The Village still gives the old zone name, Bench Maji.

T2722FORESTOTHER

• Exact match with T.272239%

• Very close, same Gesha genetic group24%

• Genetic background unrelated to Gesha37%

Figure data

Group	Count
Exact match with T.2722	39
Very close, same Gesha genetic group	24
Genetic background unrelated to Gesha	37

SourcePruvot-Woehl et al., J. AOAC Int. 103(2):325–334, 2020; WCR news, 2020-03-23 · values as publishedCited source

What is known geneticallyHow far the name can be checked

The T2722 lineTraceable to the 1953 accession number. WCR calls it uniform but gives no sourceCan be compared against T.02722

The source forest populationNever uniform within; Gesha Village grows a line specifically to keep that diversityCan be checked to the forest and the block, but does not point to any single sample

Called Gesha elsewhereThey vary. One case in Ecuador in 2026 was a different Ethiopian landraceCan only be tested case by case

SourceWCR variety catalog; Gesha Village official site; Tapia-Armijos et al. 2026; Pruvot-Woehl et al. 2020 · grouping by OLLA's editorsCited source

Geisha · Gesha · Gori Gesha

Geisha

The earliest spelling in the germplasm records, the one research archives and Panama have always used

§03

Gesha

The usual English spelling of the mountain near the collecting site; WCR says the two spellings are interchangeable

§03

Gori Gesha

A forest in the West Omo Zone, and also the name of one Gesha Village line. It is not the Gesha district in the Kafa Zone

§05

SourceWCR variety catalog; Gesha Village official site; BMC Plant Biology 2025Cited source

Section 03F3 ↑

3. The road T2722 travelled

The day it arrived

WCR records the route like this. The material was collected from Ethiopian coffee forest in the 1930s, went first to the Lyamungu research station in Tanzania, and reached Central America in 1953, where it was registered as T2722.

Anacafé, Guatemala's coffee association, gives the day of arrival in its 2020 variety guide. It was 28 July 1953, and the material was received by IICA's coffee introduction garden at Turrialba. CATIE's own publication wrote in 2020 that it entered CATIE's international coffee collection from Tanzania in 1953, using the institution's present-day name.

Further back, two chronologies

One account says 1931. Writing in Roast magazine in 2006, Willem Boot passed on the findings of Labouisse, a researcher at the French agricultural research body CIRAD. On this account the material was collected in 1931, sent to the Kitale research centre in Kenya between 1931 and 1932, and planted in 1936 at two research stations, Kawanda in Uganda and Lyamungu in Tanzania.

The other account says about 1936. Hacienda La Esmeralda's official site and the 2024 genome paper by Medrano and colleagues both put it this way, saying the material went first to Kenya and then to Tanzania. The Medrano paper cites an article Boot wrote in 2013.

In 2017 Daily Coffee News mentioned a 1936 letter from the British consulate that refers to Geisha Mountain several times. WCR says only “1930s” and does not name a collector.

The collector most often named in these retellings is Richard Whalley, the British consul for the Bench Maji area, and Esmeralda says he was instructed to collect about 10 pounds of seed. His name appears only in the line of retellings that runs through Esmeralda and Boot. We found no original document from the 1930s, and neither the consular reports nor the records of Kenya's agriculture department have been made public, so we do not state it as fact. Gesha Village named one of its lines “Gesha 1931” to commemorate that expedition, and the name is not independent evidence for the date.

To Panama

CATIE's 2020 article says it was distributed to Panama from 1963. Boot wrote in 2006 that it was Francisco “Don Pachi” Serracín who brought it into Panama in 1963. In 2017 Daily Coffee News quoted Price Peterson, who also named Serracín.

WCR says “1960s”, and Medrano 2024 says the mid-1960s. We write “from 1963”, and following Boot and Peterson we record the person who introduced it as Serracín.

Why it was planted, and why it was then neglected

WCR says it was distributed to Panama via CATIE “after it had been recognized for tolerance to coffee leaf rust”. But leaf rust was first reported in the Americas only in 1970, in Brazil, and it spread through Central and South America in the 1970s and 1980s.

So any claim in the 1950s or 1960s that it tolerated rust can only have come from tests in a germplasm garden, an experimental plot kept to conserve and test planting material. Panama's coffee farms had no leaf rust at the time.

Esmeralda tells it differently again. Its official site says that what actually brought the seed to the farm was Gesha's resistance to ojo de gallo, a leaf-spot disease of the Americas. Among the sources we read for this issue, none says it was introduced back then because it tasted good.

Why nobody went on to grow it is also on record. WCR writes that its branches are brittle and farmers did not like it, so it was never planted widely. In a 2014 bulletin the Costa Rican agronomist Ramírez wrote that multi-site trials from 1963 onwards followed it over 7 harvests, that its yield showed no clear difference from the Typica control, and that the conclusion in the 1970s was not to recommend it unless the price made up for the low yield. This rests on that one source only.

WCR also changed its ratings in October 2025. Yield potential went from Medium to Low, and leaf rust from Tolerant to Intermediate resistance. The leaf rust change may simply be a site-wide renaming of the label, so it cannot be read as saying that Gesha has become less resistant.

1930sKENYALYAMUNGUTURRIALBA 1953PANAMA 1963

1931The 1931 account

Boot, writing in Roast in 2006, passing on CIRAD's Labouisse. On this account it was collected in 1931, sent to Kitale in Kenya between 1931 and 1932, and planted at Kawanda in Uganda and Lyamungu in Tanzania in 1936.

1936The about-1936 account

Given by the Esmeralda site and by the 2024 paper by Medrano et al.: first to Kenya, then to Tanzania. The collector in these retellings is the British consul Richard Whalley, but no document has been seen.

GAPWhat both accounts lack

WCR says only 1930s. Neither account can produce an original document from the 1930s.

SourceBoot, Roast 2006; Hacienda La Esmeralda official site; Medrano et al., G3 2024; Daily Coffee News 2017; WCR variety catalogCited source

ClaimSource and limits

Rust toleranceDistributed to Panama after being recognised as rust-tolerantWCR. Leaf rust was first reported in the Americas only in 1970, so this can only be a germplasm-garden test result

ojo de galloWhat brought the seed to Esmeralda was its resistance to this leaf-spot disease of the AmericasEsmeralda official site

Brittle branchesFarmers did not like it, so it was never planted widelyWCR

Ordinary yieldOver 7 harvests its yield showed no clear difference from Typica; not recommended in the 1970sRamírez 2014, the only source

Current ratingYield potential Low; leaf rust Intermediate resistanceWCR, since October 2025. The leaf rust change may only be a relabelling

SourceWCR current page and archived copies from 2023-06, 2025-06 and 2025-10; Talhinhas et al. 2017; Hacienda La Esmeralda official site; Ramírez 2014Cited source

Section 04F4 ↑

4. Planted, then discovered

It had been there all along

Gesha had been in Panama since 1963, as chapter 03 showed. In 1997 Hacienda La Esmeralda bought the land that later became the Jaramillo farm and planted Gesha on plots above 1,650 m. Its official site says this was higher than anyone had grown Gesha before. The site does not say in which year the first Gesha was planted at Jaramillo, so we cannot work out how many years passed on this farm between planting and recognition.

What happened in 2004

By Esmeralda's account, to enter that year's competition the Peterson family, for the first time, “separated production from different areas of the farm into individual lots” during processing, in the words of its official site. Until then the coffee from all the plots had been processed together.

SCAP's 2004 results table lists 11 lots in all. The highest price went to PA101, “Jaramillo Special”, at $21.00 a pound, or $46.30 a kilo. The lot was 7 bags of 60 kg each, 420 kg in total.

It sold for $19,445.16 in all, 35.51% of the auction's total of $54,755.98. The other 10 lots fetched only $1.25 to $2.53 a pound, and the top price was 8.30 times the second.

PA101 was won by The Roasterie, Public Market Coffee and Groundwork Coffee. A note under the table says they had agreed to share the lot with Intelligentsia, G.H. Howell, Stumptown and Sweet Maria's, one bag for each of the 7 companies.

The results table does not name the variety. Identifying Jaramillo Special as Esmeralda's Gesha is our judgement, based on Esmeralda's own account: it says it won in 2004 with Gesha, and Jaramillo is its farm. A 2010 retrospective gives the price as $20 a pound, and we go by SCAP's results table.

BEFORE 20042004

1. Jaramillo Special$21.00
2. Don Pachi Estate$2.53
3. Café Leru$2.30
4. Café De Eleta$2.20
5. Lerida Estate$2.15
6. Cafetales Doña Berta$2.04
7. Café Don Pepe$1.92
8. La Victoria Estate$1.90
9. Arkapal Premium$1.85
10. Gian Panamá$1.26
11. Sohel Café$1.25

Figure data

Item	Value	Rank
Jaramillo Special	21	$21.00
Don Pachi Estate	2.53	$2.53
Café Leru	2.3	$2.30
Café De Eleta	2.2	$2.20
Lerida Estate	2.15	$2.15
Cafetales Doña Berta	2.04	$2.04
Café Don Pepe	1.92	$1.92
La Victoria Estate	1.9	$1.90
Arkapal Premium	1.85	$1.85
Gian Panamá	1.26	$1.26
Sohel Café	1.25	$1.25

SourceSCAP · Best of Panama 2004 Final Results · 2004-07-03 (Wayback archive) · values as publishedCited source

Section 05F5 ↑

5. The source is messier than the market

Gesha Village's dates

Gesha Village is in south-western Ethiopia, and the home page of its official site says the West Omo Zone. The site gives no founding year. In 2011 it signed the lease on its land, as Barista Magazine and Daily Coffee News reported in 2017. In the same year the farm collected seed from the Gori Gesha forest about 20 km away, as its own site says.

The Our Story page of the site is signed by Adam Overton, Rachel Samuel and Willem Boot. The farm covers 471 hectares in all. The site gives the planted area in two places, once as 320 hectares and once as 341, and the eight blocks add up to 321.3 hectares. The altitude runs from 1,909 to 2,069 m.

The three lines

The site's variety page describes the three lines in very different terms. Gesha 1931 was selected from the diverse forest population for its resemblance to the Panama Gesha, judged on how the plants look, the shape of the beans and the cup.

Gori Gesha is meant to reproduce the genetic diversity within the Gori Gesha forest, and its seed was collected on the farm's own expedition in 2011. Illubabor Forest is a cultivated line from an Ethiopian research centre, selected from a 1974 expedition to the Illubabor coffee forest, and the site says it is disease-resistant.

So even at the source, “Gesha” does not point to one thing. One line was selected to match the Panama line, one was deliberately left unselected, and one comes from another forest.

The site also says that historical research and genetic testing have confirmed that the Gori Gesha forest where it collected seed is where the Panama Gesha was originally collected. The test data have not been published, so we record this only as the farm's claim.

Eight blocks

Each of the eight blocks grows one line. Gori Gesha is planted in Shewa-Jibabu, Shaya and Bangi, Gesha 1931 in Surma, Oma and Narsha, and Illubabor Forest in Dimma and Gaylee. These block names often turn up in the lot names at the Pride of Gesha auction, which chapter 12 draws on.

GESHA 1931GORI GESHAILLUBABOR

LINE	WHAT THE SITE SAYS
Gesha 1931	Selected from the forest population for resemblance to the Panama Gesha, judged on plant, bean shape and cup
Gori Gesha	Reproduces the genetic diversity within the Gori Gesha forest; seed from the farm's own 2011 expedition
Illubabor Forest	A cultivated line from an Ethiopian research centre, selected on a 1974 expedition to the Illubabor coffee forest; the site says it resists disease

SourceGesha Village · The VillageCited source

BLOCKS	LINE	ALTITUDE m	AREA ha
Shewa-Jibabu	Gori Gesha	1,973–2,069	48.5
Shaya	Gori Gesha	1,926–2,069	36.5
Bangi	Gori Gesha	1,911–2,001	54
Surma	Gesha 1931	1,909–2,063	45.9
Oma	Gesha 1931	1,931–2,040	67.6
Narsha	Gesha 1931	1,963–1,977	5.3
Dimma	Illubabor Forest	1,966–2,019	28.7
Gaylee	Illubabor Forest	1,916–1,982	34.8

SourceGesha Village · block table on The Village page · values as published; total calculated by OLLACited source

Section 06F6 ↑

6. What the DNA says

This chapter goes through the papers one at a time. For each, it says only what was tested and what was found, and claims nothing further on the paper's behalf.

Three papers find it uneven

The 2020 paper by Pruvot-Woehl and colleagues appeared in J. AOAC Int. They tested 2,533 samples with 8 SSR markers, and of the 88 held as Gesha, only 39% matched the reference plant T.2722 exactly, as chapter 02 described.

The 2024 paper by Scalabrin and colleagues appeared in Nature Communications. The T.02722 they obtained from CATIE has a stretch on chromosome 1 whose structure differs from the other samples, which the paper describes as a heterozygous homoeologous exchange. Neither the public Geisha data nor the Geisha sample supplied by WCR has this stretch. On this basis the authors point out that germplasm of the same name can differ structurally in this way.

The paper by Tapia-Armijos and colleagues appeared in Frontiers in Plant Science on 14 July 2026. A farm in Loja, Ecuador, had bought “Gesha” seed at $500 a kilo, and when it was tested with 11 SSR markers it proved to be a different Ethiopian landrace, similar to T.02722 but not the same.

The other two are about something else

The 2024 paper by Salojärvi and colleagues in Nature Genetics did not assemble a Gesha genome. It reaches two conclusions that bear on Gesha. One is that the recently established Geisha cultivar is admixed in both of its subgenomes and resembles about half of the wild material. The other is that the area around Gesha mountain holds a concentration of wild material suited to domestication.

In 2024 Medrano and colleagues published the first chromosome-level Gesha genome in G3, 1.03 Gb in size. The tree they sampled grows at Goodland Organics in Goleta, California, from seed that Price Peterson sent from Panama in 2006. It is the genome of one tree, and it cannot tell us whether other trees called Gesha are the same.

Two claims with no paper behind them

The WCR catalog says that recent genetic diversity analyses confirm the Panamanian line descended from T2722 is “distinct and uniform”, meaning different from other material and consistent within itself. It gives no source for these analyses.

On 20 August 2026 the EFE news agency put out a wire story that many outlets reprinted. It said SCAP had announced a comparison carried out in France, and quoted its lead, Christophe Montagnon, as saying that Panama's Gesha and CATIE's original material are genetically identical: “la genética es exactamente la misma”.

Neither the method, nor the number of samples, nor a paper has been published for this comparison. Both claims can be cited only as statements by the institutions, not as research findings.

They do not contradict each other

The 39% study tested samples that were held as Gesha, whereas WCR and SCAP are reportedly talking about Panama's T2722 line. The two tested different things, so both can be true at once.

Read together, they make the statement that Gesha is not one variety stronger. Under the one name there are at least the T2722 line, the source forest population, and other Ethiopian landraces sold as Gesha, such as “Evangelina”. That is our judgement.

Section 07F7 ↑

7. Where it grows now

The earliest documented planting

Meeting the standard here means having at least one primary source, or two secondary sources independent of each other. By that standard we can give a year for nine places.

Costa Rica comes first, in 1953, at IICA's introduction garden in Turrialba. Panama follows from 1963. Mexico's year is 1980, when the Mexican Coffee Institute, INMECAFE, introduced it. California's is 2006, when Price Peterson sent seed to Goodland Organics.

Thailand's is 2010, when Doi Chaang trial-planted 40 to 50 Ethiopian seeds. In Hawaii the seed entered the United States in 2010 and was trial-planted on Maui in 2011. In Ethiopia, Gesha Village collected its seed in 2011.

Yunnan's is 2013, when the Yunnan Academy of Agricultural Sciences brought in a small amount from Panama. The report was published in two places, and elsewhere in the text the institution is named as the Academy of Social Sciences.

For Colombia all we can fix is a lower bound, after 2008. Café Granja La Esperanza's official site says it leased La Carleida in Panama in 2007 and won first place at BOP a year later. “At this point”, the site says, came the decision to bring Geisha seed back to Colombia. The original gives no year. “After 2008” is our inference, and an earlier introduction cannot be ruled out.

Brazil and Taiwan each have only one report in the trade press, which falls short of the standard. The Brazilian one says it entered the germplasm collection of the IAC institute between 1952 and 1953, and the Taiwanese one says 2014. For Peru, Guatemala, Honduras, El Salvador, Nicaragua, Ecuador, Bolivia, Malawi, Rwanda and Burundi, Vietnam, Indonesia and Australia we found no first-planting year that meets the standard.

Malawi needs a word of its own. WCR says Geisha is “cultivated widely” there. A handbook from the Coffee Association of Malawi says the line promoted locally was a single-plant selection from “Geisha 56”, chosen for its resistance to Fusarium wilt, but it gives no year.

The altitudes, laid out

The OLLA Codex keeps its own directory of Gesha farms. Calculated from version 607 of the Codex, the directory lists 51 farms, 38 of which have an altitude range on record. They are spread across 17 countries and regions and run from 737 m to 2,500 m.

The chart includes the 10 countries that have two or more farms on record. Panama has 7, between 1,300 and 2,000 m. Gesha Village at the source sits at 1,909 to 2,069 m, and it is the horizontal band in the chart.

These numbers will be revised along with the Codex. Of the farm records used in the July chart, several have since been withdrawn from the Codex or had their altitude changed.

No ruler yet

This chart cannot be read as showing where Gesha does better. It shows the altitude ranges of farms or growing regions, not the altitude of the plots where the Gesha grows. In a region that runs from low ground to high, the Gesha may be planted only on the highest plot.

Nor were these origins cupped on a common form. Each impression comes from a different table, different judges and a different roast level. Add what chapters 02 and 06 describe, that the material introduced in each place may not be the same. That is our judgement.

Section 08F8 ↑

8. Twenty-three years of prices

How the line is drawn

Before 2017 Best of Panama had no separate Gesha category. The official results give only lot names, and in some years the category codes mix in other varieties such as Pacamara. So for 2004 to 2016 we can take only one lot a year, the highest-priced lot that year that can be shown to be Gesha.

What identifies the lot as Gesha differs from year to year. Sometimes Geisha is in the lot name, sometimes it is contemporary reporting, sometimes what the winning bidder said on the day, and sometimes the farm's own account. For the 2009 “Esmeralda Special” there is no contemporary report, and we infer that it was Gesha from the name of the series it belongs to. From 2017 there is a Gesha category, and every lot is on record.

2008 is a gap. The top price that year was for PA2008-01, “La Carleida / Café Granja La Esperanza”, at $47.00 a pound. The official results graphic does not give the variety, and the farm itself says only that it won, not what variety it was. So for 2008 we record the highest-priced lot with Geisha in its name, at $22.00 a pound. For the 2013 top price, SCAP's results graphic says $350.25 a pound and a 2018 Lamastus blog post says 350.75; we go by the results graphic.

Thirteen times

In nominal dollars, the year's top Gesha beat every earlier year in 2004, 2006, 2007, 2010, 2013, 2017, 2018, 2019, 2020, 2021, 2023, 2024 and 2025, 13 times in all. The 2026 price of $18,004 a kilo did not break the record.

Of those 13, Hacienda La Esmeralda set 7, Elida and the Lamastus family 3, and Finca Sophia, Finca Nuguo and Carmen Estate 1 each. Across all 23 editions, the year's top Gesha came from Esmeralda 10 times and from Elida and Lamastus 5 times; the other 7 came from a different farm each, and 2008 is undetermined.

The 2025 price of $30,204 a kilo is 652.4 times the $46.30 of 2004. That is a ratio of nominal prices, not adjusted for inflation.

The top lot keeps shrinking

The weight of the top lot has fallen all the way. It was 420 kg in 2004 and 226.80 kg in 2007. From 2011 to 2022 it was 100 pounds, or 45.36 kg. It was 25 kg in 2023, 20 kg in 2024 and 2025, and only 15 kg in 2026.

Its share of the whole auction's sales was 35.51% in 2004, 46.17% in 2007, 45.94% in 2010 and 21.11% in 2025. In 2026 it was only 8.88%, the lowest since the Gesha category began. The share was also low in 2011, 2012, 2014 and 2016, between 4.37% and 7.65%, but a lot was then only 100 pounds and prices were low, so those years do not compare directly with today. The 2013 and 2015 results graphics give no auction total, and we did not add the rows up ourselves, so those two years are left blank.

Since 2017 the Gesha category has taken between 86.26% and 93.64% of the auction's sales each year, and 90.41% in 2026. In other words, the Best of Panama auction today is essentially a Gesha auction. The auction totals for 2016 to 2018 are our own sums of every lot, with no official figure to check them against.

Section 09F9 ↑

9. No record in 2026, and still it cost more

Four numbers, two directions

Compare the Gesha category in 2025 and 2026. The top price fell from $30,204 to $18,004 a kilo, down 40.39%. The median rose from $1,166.50 to $3,019, 2.59 times what it had been.

The weighted average price is the category's total sales divided by its total weight in kilos. It rose from $3,258.10 to $4,583.68 a kilo, up 40.69%.

Total weight fell from 800 kg to 600 kg, because each lot went from 20 kg to 15 kg. The Gesha category's total sales rose from $2,606,480 to $2,750,205.

The lowest price went up as well. All 40 Gesha lots in 2026 sold at $1,504 a kilo or more, whereas the lowest in 2025 was $674 and the lowest in 2024 was $527.

Lots at $1,500 a kilo or more went from 19 to 40. Those at $3,000 or more went from 10 to 21, and those at $5,000 or more from 7 to 12.

More high-priced lots

There was 1 Gesha lot at $10,000 a kilo or more in 2023, 2 in each of 2024 and 2025, and 5 in 2026. Of the 5 in 2026, 2 came from Elida, 2 from Hacienda La Esmeralda and 1 from Lérida. The highest was Elida's GW-01, at $18,004 a kilo.

The longer view

The Gesha category's median was $115.74 a kilo in 2017, $206.69 in 2020, $671 in 2023, $1,008 in 2024, $1,166.50 in 2025 and $3,019 in 2026. In the ten years with lot-by-lot data, the median line has never once fallen back.

Over the same ten years the top-price line fell back twice, in 2022 and in 2026, and chapter 08 has the chart. The record is one lot; the market is all the others.

These are facts about the distribution, not causes. The weight per lot and the total weight changed at the same time, and so did the buyers. No source says that less supply pushed the price per kilo up, so we do not claim cause and effect.

The way processing is written is changing too. In 2025 there were only two labels, Washed and Natural, with 20 lots each. In 2026 there were 19 Washed, 8 Natural, 9 “Natural + Fermentación (o Natural Anaeróbico)”, 3 under two labels beginning with Honey, and 1 Natural CWF. We record them as written and do not group them.

Section 10F10 ↑

10. Scores and prices

The score is the cupping score the competition judges give each lot. In 2026, for example, the competition results were announced on 3 August and the auction did not open until 23 September, so buyers knew the scores when they bid.

Correlation, year by year

A correlation coefficient measures how closely two sets of numbers rise and fall together. 1 means they move in perfect step, and 0 means there is no relation. We calculated the Pearson correlation, within the Gesha category, between the score and the logarithm of the price per kilo. We take the logarithm because the prices differ by several orders of magnitude.

The coefficient was 0.675 in 2019, a year with 36 lots. In the following years it was 0.755 (2021), 0.707 (2022), 0.835 (2023) and 0.828 (2024). It rose to 0.930 in 2025 and was 0.905 in 2026, with 40 lots in each of those two years. The Spearman correlation, calculated on ranks, was 0.611 in 2019 and 0.849 in 2026.

In 2020 the platform published scores only for the top 20 lots, so the resulting 0.876 is listed separately and not compared with other years. The 2017 and 2018 results pages have no scores.

Top score and top price

From 2019 to 2023, and again in 2025 and 2026, the highest-scoring lot in the Gesha category was also the highest-priced. 2024 was the exception.

That year the top score went to GW-01 Kaizen, with 95.50 points, at $10,002 a kilo. The top price went to GN-01 Elida Geisha Natural Vuelta, with 95.13 points, at $10,013 a kilo. The two were only $11 apart.

What it can and cannot say

In these years prices have tracked scores more closely than in 2019 to 2022, but that only means the rankings within the category line up. It does not show that high scores caused high prices, because lot weights, categories and buyers all differ from year to year.

Nor can it be used to compare price levels across years. The lowest-scoring lot of 2026, at 92.13 points, sold for $1,554 a kilo. Of the 9 lots that scored 93 or more in 2021, 7 went for less than $800 a kilo.

What has changed is the price level of the whole auction, not the scores. That is our judgement.

Section 11F11 ↑

11. Who is buying

The two first places of 2018

This chapter begins by correcting a mistaken inference in the July edition. Daily Coffee News's report of 20 July 2018 gives $803 a pound in its headline, but its text says Black Gold bought the Green Tip Natural at “$830 per pound”. The report also says that the $661 a pound Saza Coffee paid was for a split of the same coffee. From this the July edition inferred that 803 was the average of 830 and 661, weighted by weight.

The official results page records it differently. The $803 a pound was for GN-1, “ELIDA GEISHA GREEN TIP NATURAL”, a natural. The $661 a pound was for a different lot, GW-1, “ELIDA GEISHA GREEN TIP WASHED”, a washed coffee. Each lot was 100 pounds, and the page shows no sale at $830 a pound.

Lamastus's own 2018 blog also lists the washed and the natural as two first places. So 661 was the price of the washed lot, not the price of a split of the natural, and the “weighted price” inference is withdrawn.

The same name

In the 15 editions from 2010 to 2024, Saza Coffee was on the winners' list for the year's top Gesha every time except 2014, 14 editions in all. In 2018 it appears as “Sourced for: SAZA COFFEE”. In 2011, 2012, 2013 and 2017 the lists themselves add “Japan”. The 2025 winner was Julith Coffee and the 2026 winner was JD.com, and neither list includes Saza.

One name, or a string of names

Sorting the winners' lists for the top Gesha of all 23 editions by how many names they carry, 8 have a single name: 2005, 2008, 2009, 2010, 2014, 2016, 2025 and 2026. A proxy purchase written as “Wataru for” a company counts as one name.

The other 15 have two or more names, and that includes all eight editions from 2017 to 2024. The 2019 list has 13 names, and the 2020 one is longer still. By 2025 and 2026 the top lot was back to a single name each year.

The number of names is not the number of companies paying. No edition has ever disclosed how joint winners divided the coffee or shared the cost, so we do not write that the top lots are all shared purchases.

NATURALWASHED$803/LB$661/LB2018 BOP

Daily Coffee News 2018-07-20Official results page

USD 803The record price in the headlineThe price of GN-1, the natural lot

USD 830The text says Black Gold bought the natural at this priceNo such sale on the page

USD 661Said to be Saza's share of the same coffeeThe price of GW-1, the washed lot, which is a different lot

Weighted priceInferred from this in the July edition, withdrawn in this oneDoes not hold

SourceDaily Coffee News, Nick Brown, 2018-07-20; Best of Panama Auction 2018Cited source

2004PA101 Jaramillo Special

The Roasterie, inc; Public Market Coffee Co.; Groundwork Coffee Co.

NAMES

Several

2005pa052505-01 Hacienda La Esmeralda

Volcafe Specialty Coffee

NAMES

One

200601-CC2115 Hacienda la Esmeralda

Small Axe Coffee Alliance (Sweet Maria's Coffee Inc, Stumptown Coffee Roasters, Intelligentsia Coffee Roasters, Groundwork Coffee Company and Kaffa)

NAMES

Several

2007pa2007-01 Hacienda la Esmeralda

The Roasterie, inc.; Intelligentsia; Roastmasters.com (Willoughby's Coffee & Tea); Zoka Coffee Roaster; 49th Parallel Roasters; Groundwork Coffee Co.; Coffee Klatch Roasting

NAMES

Several

2008PA2008-06 Geisha Aristar Agrario La Valentina

Wataru for Gabee Yang

NAMES

One

2009PA09-01 Esmeralda Special

Wataru for Ogawa Coffee

NAMES

One

2010PA-01 Esmeralda Special (1217)

Saza Coffee

NAMES

One

Saza

On the list

2011PA-NAT-A Don Pachi Geisha

Wataru for BACH KAFFEE (Japan), SAZA COFFEE (Japan)

NAMES

Several

Saza

On the list

2012PA-N-01 Kotowa Don K Geisha Natural

Horiguchi Coffee & SAZA COFFEE (JAPAN)

NAMES

Several

Saza

On the list

2013PA-N-01 Esmeralda Special Natural C.V.

Saza Coffee (JAPAN); HAAYA GOURMET COFFEE CO., LTD (TAIWAN)

NAMES

Several

Saza

On the list

2014PA-GN-01 La Mula Natural CAC

Junior Coffee

NAMES

One

2015PA-EN-01 Esmeralda Geisha Natural Jaramillo

Saza Coffee & HAAYA GOURMET COFFEE CO., LTD.

NAMES

Several

Saza

On the list

2016PA-EW-01 Elida Geisha Green Tip Washed

Saza Coffee

NAMES

One

Saza

On the list

2017GN-01 Esmeralda Geisha Cañas Verdes Natural

KEW Specialty Coffee Ltd, Sydney Coffee Business Pty Ltd, SAZA COFFEE JAPAN

NAMES

Several

Saza

On the list

2018GN-1 ELIDA GEISHA GREEN TIP NATURAL

BLACK GOLD COFFEE CO. LTD & OKLAO INTERNATIONAL CO. LTD. & CHANG WEI SHIN Sourced for: SAZA COFFEE / KLATCH COFFEE/COFFEE TREASURES/SHUNDA FOOD FLAVOR & INGREDIENTS CO., LTD

NAMES

Several

Saza

On the list

2019GNEP-01 ELIDA GEISHA GREEN TIP NATURAL

SAZA COFFEE, Haaya Gourmet Coffee Ltd., Klatch Coffee USA, Black Gold Coffee Co. Ltd. (Taiwan), OKLAO Specialty Coffee, Grand Cru Coffee Ltd (China), Willoughby's Coffee & Tea, Aroma Coffee Australia, KEW SPECIALTY COFFEE for COFFEE TECH LTD(coffee-tech.co.nz), HWC ROASTERS FOOD AND BEVERAGE CO. LTD., The Hub Coffee Malaysia, Difference Coffee Co., Dragonfly Coffee Rosters

NAMES

Several

Saza

On the list

2020GW-01 FINCA SOPHIA OLYMPUS GEISHA LAVADO – C460

PICK COFFEE 立裴米緹 / 上海咖威Cooway&音季 / S.O.E Coffee / June’s Garden Coffee六月花园咖啡 / CMLab罗锦铭 / 致品山下 / Liming利茗精密減速機 / GK咖啡x平安居咖啡民宿 / TUMAZ Coffee黑熊咖啡 / SAZA COFFEE / BlackDrum Roasters / kafeD / Yi Zheng Coffee Roastery 昆山亿政 / C•T台北uSmile coffee北京易大 / Ningbo AncientCoffee宁波愛咖啡 /

NAMES

Several

Saza

On the list

2021GN-01 Nuguo Fermented

SAZA COFFEE Grand Cru Coffee (China) Dragonfly Coffee Roasters DIFFERENCE COFFEE Coffee-Tech.co.nz / BC Co. NZ HWC Roasters

NAMES

Several

Saza

On the list

2022GN-01 Guarumo Coffee Farm Black Jaguar Natural Limited

SAZA COFFEE ・Grand Cru Coffee(China) ・Coffee-Tech.co.nz / L'escargot d' or cosmetocafe ・ECOFARM COFFEE 意可坊咖啡 ・Black Gold Specialty Coffee / Oklao Specialty Coffee / The Espresso Lab ・杭州资董窝玺&董咖啡 ・HWC Coffee ・DIFFERENCE COFFEE ・Fisher coffee 啡舍, ・东方国际商业(集团)有限公司, ・珈露梦咖啡(carmo), ・Rondo Cafe, Proud Mary Coffee

NAMES

Several

Saza

On the list

2023GW-01 Carmen Geisha - Carmen Estates By Panama Red Carmen Café Trading

Coffee Tech Limited (NZ)/coffee-tech.co.nz ; SAZA COFFEE ; 格米莱GEMILAI /Corrima coffee machine ; CHG & MoveRiverCoffee.公子高.lzkngy ; Difference Coffee ; L'escargot d'or cosmetocafe, The Espresso Lab, Grand Cru Coffee (China), 叁生恬铺

NAMES

Several

Saza

On the list

2024GN-01 Elida Geisha Natural Vuelta

SAZA COFFEE, Grand Cru Coffee (China), OKLAO SPECIALTY COFFEE, BREW BOYY CREW, Ruber.Roastery(Korea), Pebble Coffee for Simple Kaffa, Paradise Coffee Roasters, E.C, 制糖师, 叁生恬铺杭州旺焙学院, 乜早, One by One Cafe (Tokyo & Taizhou), CHG手多多X Pincle拼一口咖啡

NAMES

Several

Saza

On the list

2025GW-01 Hacienda La Esmeralda Geisha Washed

Julith Coffee

NAMES

One

2026GW-01 Elida Geisha Torre

京东 JD.com

NAMES

One

SourceSCAP results graphics and pages, all years; Best of Panama auction platform · names counted by OLLA from the listsCited source

Section 12F12 ↑

12. Gesha at other auctions

Two private auctions in Panama

The earliest documented edition of Hacienda La Esmeralda's own auction is the “Esmeralda Special Auction” of 22 May 2008. An auction lot was then 300 pounds, and the top price in 2008 was $105.25 a pound. The results pages for 2008 to 2011 do not give the variety.

From 2022 it switched to pricing per kilo, with most lots at 10 kg. The top price was $7,532 a kilo in 2024 and $14,964 a kilo in 2025, though that lot was only 0.8 kg. In 2026 it was $10,253 a kilo, for a 10 kg lot.

Lamastus Family Estates also held a private auction on 9 September 2026. The top lot was Luito Geisha Washed L1, at $14,710 a kilo, and it was only 2 kg.

The auction at the source, Pride of Gesha

Gesha Village's auction was called the Gesha Village Coffee Auction in 2017 and 2018 and has been Pride of Gesha since 2019, and by 2026 it had been held ten times. The top price rose from $85.10 a pound in 2017 to $735 a pound in 2026, or $1,620.40 a kilo. The 2026 lot was GVA.RSV.5, of the Gesha 1931 line, washed and dried in a dark room.

It was only 19.84 pounds, or 9 kg, whereas the top lots of earlier editions were all between 132 and 198 pounds. The price went up and the lot got smaller, and anyone citing this record should say both.

A trade publication said in 2022 that Pride of Gesha's $254 a pound that year was the highest price in the history of African coffee. That is no longer true. In 2025 and 2026, Ethiopian lots that were not Gesha sold for $1,604 to $2,376.50 a kilo. The two highest Ethiopian auction prices we saw in 2026 were ALO's $2,275 a kilo and Faysel Abdosh's $2,376.50 a kilo, and neither lot was Gesha.

Cup of Excellence

The Cup of Excellence (CoE) is a competition and auction held separately in each producing country and run by the Alliance for Coffee Excellence (ACE). For the years that can be found on the ACE website, the highest CoE price in seven countries, Guatemala, Costa Rica, Brazil, Nicaragua, Honduras, Colombia and El Salvador, was set by a Gesha lot.

Ecuador's top price was for the variety Sydra. For Mexico and Peru, the page for the year of the top price has no variety column, so we cannot tell.

The highest we saw was El Injerto I's natural Gesha in Guatemala in 2026. Split lot 1C sold for $449.10 a pound, or $990.10 a kilo, and weighed only 33.1 pounds. The other two lots of the same coffee, 219.3 pounds each, sold for only $161.30 and $161.10 a pound.

For years with a single national ranking, we counted the lots containing Gesha in the CoE top ten. Guatemala 2023 and Peru 2023 each had 10, Peru 2024 had 9, Costa Rica 2023 had 8, Colombia 2023 and Honduras 2024 each had 7, and Nicaragua 2024 had 3.

Lots whose variety column lists a mix that includes Geisha are counted too. From 2025 several countries switched to grouping lots by processing method, so the top ten was no longer one ranking and we stopped counting.

$30,204 is still the highest

As of 26 September 2026, within the scope of our search, we found no coffee sale on public record above $30,204 a kilo. The search covered the auction platform pages of Best of Panama, Esmeralda, Lamastus, Pride of Gesha, ALO and Faysel Abdosh, 197 CoE results pages across 11 countries, the main trade media, and searches in Chinese, English and Japanese. Private negotiations whose results are not published, and platforms inside China, were outside that scope.

Section 13F13 ↑

13. What an auction measures, and the layers in the price

What a multiple says

On 7 August 2025, Best of Panama's GW-01 sold for $30,204 a kilo. The lot was 20 kg, scored 98 points, drew 549 bids and sold for $604,080 in all, and the winner was Julith Coffee.

The commodity coffee price means the reference price on the futures market. Fresh Cup's report quoted a commodity price of $2.98 a pound for the same day, about $6.57 a kilo. The two prices are about 4,597 times apart.

A price 4,597 times higher does not make the coffee 4,597 times better. The price was bid up by buyers from many countries on a lot of only 20 kg, and it contains scarcity, publicity value and the buyers' own marketing budgets.

The two prices were never the same kind of thing. One is the sale price of a small lot at auction, the other a reference price on the futures market. Reading the gap as a scale of quality mistakes an auction for a cupping. That is our judgement.

A different yardstick

The same record price is 18.64 times the highest-priced Gesha we saw in Ethiopia, a 9 kg lot at Pride of Gesha in 2026 that sold for $1,620.40 a kilo. Both are the top prices at their own auctions, so the comparison shows where each stands at its own auction, not absolute quality across auctions.

Six layers of price

Why it costs so much cannot be answered with one number, because the price is not one cost but six premiums of different kinds stacked together. From the bottom up they are farming cost, selection cost, cup quality, identity assets, institutional endorsement, and publicity and collecting.

Identity assets are what the name and the provenance are worth in themselves, and institutional endorsement is the credibility that competitions and auctions confer. The lower three layers are facts of farming and quality control, the upper three are market and story, and the higher you go the harder they are to verify. The only layer that can be checked by blind tasting on the cupping table is cup quality.

20 KG

$30,204 / KGTop at BOP 2025ESMERALDA washed

4,597×Commodity coffee$6.57/KG

18.64×Top Ethiopian Gesha$1,620.40/KG

SourceBest of Panama auction platform 2025; Fresh Cup 2025-08-12; Pride of Gesha 2026 results page · multiples calculated by OLLACited source

Section 14F14 ↑

14. From the kilo to the cup

How it is worked out

Every number in this chapter is our own calculation, not one a source gives. If a cup uses 15 g of roasted coffee and roasting loses 15% of the weight, the green coffee needed for a cup is 15 divided by 0.85, about 17.6471 g.

The auction cost of the green coffee in a cup is the price per kilo multiplied by 17.6471 and divided by 1,000, the same method as in OLLA LAB's Best of Panama 2026 issue. On these assumptions, the green coffee in a cup of the year's top Gesha cost $0.82 in 2004, $5.06 in 2007, $13.63 in 2013, $23.38 in 2017, $99.91 in 2021, $176.56 in 2023, $533.01 in 2025 and $317.72 in 2026.

In the 2026 Gesha category, the median lot comes to $53.28 a cup and the cheapest lot to $26.54. In 2025 the median lot was $20.59 and the cheapest $11.89. By the same method, the commodity price on 7 August 2025 comes to about $0.12 of green coffee per cup.

The grams per cup and the weight loss are our own choices, and with other assumptions the results change in proportion. These numbers stop at the auctioned green coffee. Freight, insurance, duties, the labour of roasting and packing, and a shop's costs and profit have no public data and are not included.

One retail price

On 26 September 2026 Julith Coffee, the winner of GW-01 in 2025, was selling a 15 g pack of “Best of Panama Nido 7 – Hacienda La Esmeralda” in its online shop at 3,600.00 dirhams, marked as an auction lot. At the UAE central bank's pegged rate that is about $980, roughly 1.8 times the auction cost of the green coffee per cup by the same method.

The product page carries no lot number, and matching it to the 2025 GW-01 is our judgement. The multiple describes only the gap in price; there is no public data on how much of it is freight, tax, roasting and profit. For the 2026 lots we have not yet found a retail page from any buyer.

Section 15F15 ↑

15. Common misreadings

This chapter does not argue anything afresh. It puts a few common claims next to the evidence that corrects them, and each points back to the chapter where it is set out in full. The first two are our judgement, since no source has given a standard answer, and the rest have data behind them.

“If it's floral, it's Gesha.” Floral is a descriptor that many Ethiopian-derived materials and many processing methods can bring out. What makes Gesha recognisable is whether the aroma and the structure in the mouth hold each other up, not any single descriptor. That is our judgement.

“The heavier the fermentation, the more complex the coffee.” Heavy fermentation covers up what the variety itself expresses, and the heavier the processing, the harder it is to tell what variety lies underneath. That too is our judgement.

“The record price is the year's market.” The top price in 2026 was 40.39% below 2025's, while the median was 2.59 times 2025's. The 2026 top lot accounted for only 8.88% of the auction's sales. See chapters 08 and 09.

“If the bag says Gesha, it's T2722.” Of 88 samples held as Gesha, only 39% matched the reference plant exactly. In a 2026 case in Ecuador, “Gesha” seed tested as a different Ethiopian landrace. See chapters 02 and 06.

“It went into CATIE in 1953.” CATIE was not founded until 1973. In 1953 the material was received by IICA's introduction garden at Turrialba, a collection that CATIE looks after today. See chapter 03.

“It was introduced to Panama for leaf rust resistance.” Leaf rust was first reported in the Americas only in 1970, and the claim that it tolerated rust rests on tests in a germplasm garden. What Esmeralda remembers is its resistance to ojo de gallo. See chapter 03.

“ALO's $1,739 a kilo is the top price for Ethiopian Gesha.” That lot's variety was 74158, not Gesha. The highest price for Ethiopian Gesha that we have seen is Pride of Gesha's $1,620.40 a kilo in 2026. See chapters 12 and 13.

How to tell: four axes

Rather than counting how many aroma words a cup hits, we suggest breaking “it smells good” into four axes and noting each separately: aroma intensity, the shape of the acidity, the structure in the mouth and the length of the finish. Then see whether the four point the same way.

To use it, taste blind first and write one sentence for each axis. Only then look at the label, and see which impressions hold up and which were only expectation. After that, change one variable and compare again, such as the same farm with a different processing method, or the same beans at a different roast level.

FLORALFERMENTPRICE

What people sayWhy it does not hold

FloralFloral means GeshaMany Ethiopian-derived materials and processing methods give floral notes; what matters is whether aroma and structure hold each other up. Editorial judgement

FermentationThe heavier the fermentation, the more complexHeavy fermentation covers up the variety. Editorial judgement

RecordThe record price is the year's marketIn 2026 the top price fell 40.39%, yet the median was 2.59 times 2025's (§09)

LabelIt says Gesha, so it's T2722Only 39% of 88 samples matched exactly, and one case in Ecuador was a different landrace (§02, §06)

1953It went into CATIE in 1953CATIE was founded in 1973; in 1953 it went to IICA's Turrialba introduction garden (§03)

RustIntroduced to Panama to resist leaf rustLeaf rust was first reported in the Americas in 1970; Esmeralda remembers ojo de gallo (§03)

EthiopiaALO's 1,739 is the top Ethiopian GeshaThat lot was 74158. The highest Ethiopian Gesha price we have seen is $1,620.40 a kilo (§12, §13)

SourceSources for each row are in the chapter it points to · the first two rows are OLLA's editorial judgementEditorial

Section 16F16 ↑

16. What to ask before buying

This chapter is our advice, not something a source says. After each question we note which chapter it draws on.

Ask about the line first, whether it is the T2722 line, material from the source forest, or material that some place introduced on its own (chapters 02 and 05). Then ask about provenance, whether the seller can say how it was introduced rather than offering only the one word “Gesha” (chapters 03 and 07).

Then ask about the lot, whether it is a regular lot or an auction lot, since the two are expensive for different reasons. An auction lot should match a lot number and a winning bidder. Even the 15 g pack sold by the winner of the 2025 record lot has no lot number on its product page (chapters 11 and 14).

Then ask about processing, and whether it is heavy enough to cover up the variety (chapter 15). The last question is new in this edition. It asks whether there has been a DNA comparison, or at least a source for the introduction that can be checked. The farm in Ecuador that paid $500 a kilo for “Gesha” seed found out it was different material only after testing (chapter 06).

Ordinary buyers cannot get hold of DNA reports, and that is the reality. So the practical form of the fifth question is to ask the seller where the seed or seedlings came from, and who can vouch for it.

What you wantWhat to buy

Get to know itFirst GeshaA regular lot that states the farm and the line, washed

CompareTo tell what the variety itself expressesSame farm and same line, one washed and one natural

See the sourceTo know what the source is likeA lot from the Ethiopian source (chapter 05), tasted against a Panamanian one

SourceOLLA editorial judgementEditorial

LINE

Is it the T2722 line, material from the source forest, or something introduced locally?

PROVENANCE

Can the seller explain how it was introduced, or is there only the word “Gesha”?

LOT

A regular lot or an auction lot? If it is an auction lot, do the lot number and the winner match?

PROCESSING

Is the processing heavy enough to cover up the variety?

TESTING

Has there been a DNA comparison? At the least, where did the seed or seedlings come from, and who can vouch for it?

SourceOLLA editorial judgement, based on chapters 02, 03, 06, 11 and 14Editorial