Higher-grown coffee usually does cup better, and the number of meters is not what is doing the work. When a team in Ethiopia measured twenty lots taken from 850 up to 2,775 meters, sucrose in the green seed ran from 5.8 grams per hundred at the bottom to 25.7 at the top, so the high end is more than four times the low one, in the single compound the roast has the most use for. Chlorogenic acid, which the cup reads as harshness, moved the other way. The seeds got heavier. Caffeine did not move at all. Underneath every one of those numbers sits temperature: a seed filling on a cool mountain takes longer to finish, and a longer fill leaves more sugar in the seed. That is also why the same height means two different things in two different countries, and why the grade stamped on a bag describes where a farm sits rather than how its coffee tastes.
What does growing higher change inside the coffee seed?
The clearest measurements I have found come from Urugo and colleagues, writing in Beverage Plant Research in 2024. They took twenty Ethiopian green samples across a 1,925-meter spread and ran the chemistry against both the height of the farm and its mean temperature.
Sucrose is the one that matters most to a roaster, because it is the reservoir that browning draws on. Across those twenty samples it averaged 12.371 grams per hundred, with a standard deviation of 3.973, which is a polite way of saying the lots were nothing like each other. The lowest sat at 5.8 grams per hundred on a farm at 1,025 meters. The highest, 25.7, came off a farm at 2,775.
Three other things moved with it. Trigonelline, which breaks down in the roast into some of the compounds that make coffee smell like coffee, ran from 0.48 to 2.31 grams per hundred and tracked the same direction. Crude fat went from 10.35 to 16.83 percent. Hundred-bean weight went from 13.22 to 17.28 grams, which is the density claim with an actual number attached to it.
Chlorogenic acid went the opposite way, 8.2 grams per hundred down to 3.7. That family of acids is what heat pulls apart into guaiacol and its relatives, and those taste phenolic and burnt, so a seed carrying less of it starts the roast with less of a problem. We followed that route in more detail when we looked at where coffee aroma actually comes from.
And caffeine did not correlate with height at all. Not weakly. Not in either direction. If you have been told that mountain coffee is stronger, that particular sentence has no support in this dataset.
If temperature is the cause, why does elevation predict the chemistry better?
Perfect Daily Grind put the mechanism into the trade's hands in January 2018, and Tanya Newton stated it about as plainly as it can be stated: higher elevations go with a sweeter and more complex cup, but "it's actually correlation, not causation. The real cause of this better-quality, more delicious coffee is temperature." That is their line, it has been public for eight years, and everything below is built on top of it rather than around it.
Here is the part that is harder to say. In the Urugo data, where both variables were measured against the same twenty samples, height out-predicted temperature on every single compound. Sucrose correlated with height at R = 0.91 and with temperature at R = −0.76. Chlorogenic acid: −0.94 against height, +0.57 against temperature. Trigonelline: 0.92 and −0.62. Crude fat: 0.87 and −0.44. Seed weight: 0.86 and −0.49.
That is not evidence that meters cause sugar. It is evidence that inside one mountain range the two things are welded together and the cruder instrument happened to read cleaner. A number on a topographic map has no error bars. A mean temperature has a weather station, a sensor, a siting decision and a year of variation in it.
How welded? The systematic review that our article on Brazilian coffee leans on, by Ahmed, Brinkley, Smith and colleagues in Frontiers in Plant Science, covering seventy-three studies, reports from one of its included papers "a mean annual temperature difference of 3.2 ± 0.7°C between the high and low-altitude locations." Three degrees, give or take. That is the whole of it. And the review is explicit about why it looked at height in the first place: because "different elevations are associated with different mean annual temperatures."
The same review is worth quoting against itself, because it does not hand anyone a clean story. Twelve of its studies found that growing higher went with better sensory scores. On the chemistry, though, "the evidence regarding the effect of cultivating coffee at higher altitudes on metabolites was mixed." And on temperature measured directly, four studies found warmer growing improved the sensory result while three found it hurt. If temperature were a clean master variable, that split would not be four to three.
One more thing to be careful about, and I would rather say it than let it pass. Every source I read states that slower ripening concentrates sugar, and not one of them measures the concentrating. What is measured is the endpoint: a seed off a cool farm has more sucrose in it. The step in between, the slow fill, is reconstructed from earlier work rather than watched. It is very probably right. It is not the same kind of fact as the 25.7.
What explained coffee quality better than either of them?
Soil.
Getachew and colleagues, in Agronomy for Sustainable Development in 2022, went at this the hard way: 53 farms, 159 trees, southwest Ethiopia, between 1,500 and 2,160 meters, with temperature loggers buried ten centimeters down and reading every three hours from February to December. Their finding on temperature is squarely in line with everything above. "Green bean quality was negatively correlated to soil temperatures," and the cool sites produced the better preliminary quality.
Then they partitioned the variance. Total specialty quality: 84 percent of the variation was explained by soil chemical variables. Elevation's own marginal contribution was 3.8 percent.
Eighty-four against under four. On 53 farms in one region, what the tree was standing in mattered more than twenty times as much as how high it was standing. That is a number worth sitting with, because it reframes the whole argument. Replacing "it is the height" with "it is the temperature" is a better answer and still not the biggest one.
I will also note the detail in that paper that tells you most about why this question stays unsettled. They meant to log air temperature too. The loggers were stolen. So the best-controlled field study I can point you at is missing the variable everybody most wants, for a reason that has nothing to do with science.
What does SHB or SHG on a bag actually certify?
It certifies a farm's elevation, and in several producing countries that is all it certifies.
Barista Hustle's coffee-buying curriculum is blunt about the Guatemalan system. The country sorts coffee into "nine categories according to the elevation at which it was grown", running from "Fair" up to "Fancy Strictly Hard," and then comes the sentence that matters most: "quality and size are not a part of these classifications." They add that in 2013 Guatemala told the International Coffee Organisation it did not maintain quality standards relating to coffee at all. Meanwhile 81.4 percent of Guatemalan exports carry the top grade, which tells you roughly what a top grade held by four fifths of a crop is worth as a sorting signal.
The ICO's own market development material hedges the link carefully rather than asserting it, "while there are many exceptions, coffee beans grown at higher elevations tend to be denser, larger, and have better flavor", and it keeps the hundred-point cupping scale in a separate section, as a different system measuring a different thing. Importers and roasters are the ones who collapse the two, and you can watch it happen: one importer's glossary explains that higher-grown beans "have a greater density, and thus a better specialty cup," which is two inferences chained together and sold as a definition.
Then there is the problem of where the line sits. I went looking for one number and came back with four. For Guatemalan Strictly Hard Bean, the sources I read put the threshold at 1,300 meters, at 1,350, at 4,500 feet (1,372), and at 5,000 feet (1,524). For Honduran Strictly High Grown, three sources give 1,200, 1,350 and 1,500. One importer lists the Honduran grades and declines to attach a band to any of them, which is the most honest treatment of the three. Another well-known importer's glossary renders the threshold as "1200 feet/4000 meters," which is inverted and describes a mountain that does not exist.
The grading authority's own public site does not settle it. Anacafé's coffee types page has a section headed Altitude, and the grade names and their bands live inside a picture on that page rather than in text a reader can copy. Eight regional profiles sit one click away with no elevation figures and no classifications attached to any of them.
So a grade that four sellers define four different ways, on a threshold the issuing body publishes as an image, is being read by buyers as a quality mark that the issuing body's own curriculum says it is not. That is a lot of weight for three letters.
Does higher always mean better?
No, and the places it stops are measurable.
In the Gayo highlands of Indonesia, a 2024 study sorted farms into three bands and weighed the good beans out of a fixed sample. The middle band, 1,200 to 1,400 meters, produced the most at 2,173.8 grams. The top band, 1,400 to 1,600, came in lower at 2,013.0. The bottom band trailed both at 1,568.3. Defects, interestingly, kept improving all the way up, 14.30 percent then 10.19 then 8.50, so the two measures point in different directions inside the same dataset.
In Pu'er, across five bands from 930 to 1,520 meters, linoleic, palmitic and stearic acid all fell significantly between 1,370 and 1,520. Caffeine was highest at 1,100 meters and lowest at the top. Every sample still cupped above 81 points, which is the useful caveat: a declining number is not the same as a bad coffee.
In Cusco, Peru, a 2025 study across 1,600, 1,800 and 2,100 meters found pest damage falling as the farms climbed and secondary defects rising in both varieties it tracked. It also found that the best height depended on which tree you had planted. Typica did best above 1,800 meters, and Catimor did best at 2,100. There is no single correct elevation, because the plant has an opinion about it.
A 2026 review pulls the shape together. The strongest cup qualities cluster around 1,900 to 2,000 meters, and above that physiological stress starts taking quality back. The same review says the thing I would want a reader to leave with: elevation on its own is an insufficient predictor of quality and has to be read inside the local ecology and the farm's management.
What do Kenya and Ethiopia grade instead?
Neither of them grades by height, and the contrast with Central America is sharper than it looks.
Kenya grades by screen size and defect. AA is what passes an 18 screen and is held by a 17. AB is a mix of 16s and 17s, it is the most common grade in the country, and it is routinely excellent. PB is peaberry, a single seed where two should have grown, and it screens smaller by definition. The grade is a description of the beans in the bag, and green importers working that origin are clear that it does not tell you how the coffee cups. Kenyan coffee then goes to auction, where the buyers cup it, which is where the quality judgment actually happens.
Ethiopia goes further. Chris Kornman at Royal Coffee describes a system that combines a raw value, covering defects, shape and makeup, color and odor, with a cup quality value covering cleanliness, acidity, body and flavor. The cup is inside the grade. So the origin the trade spent years calling unstandardized grades the thing that matters, and the origin with the famous acronym grades the map.
If you want to see what a cup-based judgment looks like written down, that is what a cupping score is, and it is the number we print on every bag we sell.
What do the elevations on our own coffees tell you?
Less than I would like, and I would rather show you that than tidy it up.
We publish a cup score on every coffee on the shelf. We publish what we paid for the green on most of them. We publish no growing elevation on any of them, because there is no altitude field on a single product in our catalog. The height lives on the region page behind the coffee, which is a page the product itself does not link to, so the number a buyer is told to care about is the one we make hardest to find.
Here is what happens when you go and find it. Kenya Kiambu scored 91.5, and our page for that county puts its farms between 1,600 and 1,900 meters. Our Organic Peru Decaf scored 88.4, and San Martín runs from roughly 1,000 to 2,000. El Salvador Finca San Luis scored 86.8, and Ahuachapán runs from roughly 1,000 to over 1,800.
Five points separate the top and the bottom of that list, and the elevation bands behind them overlap almost completely. Both of the lower-scoring coffees grow in regions whose published range reaches higher than Kiambu's floor and nearly as high as its ceiling. On our own shelf, the band tells you close to nothing about which coffee scored better.
Three coffees is not evidence of anything, and I will not pretend otherwise. One of them is a decaf, and decaffeination moves a cup score for reasons that have nothing to do with a mountain. Two of the three are natural-process and one is washed. And the bands themselves are 300 to 1,000 meters wide, because they describe a region rather than a farm, which is wide enough to hide the entire effect the research measures. That last point is the honest one. Our data is not bad evidence against the mechanism, it is not evidence at all, and the reason is that we publish a region where a farm belongs.
Which is the practical version of everything above. The elevation on a bag is a real fact about a place and a weak fact about a coffee, and the only instrument that settles it is the one Ethiopia already grades with and Kenya's auction already runs on. Taste it.
Questions we get about elevation and coffee
Does high-altitude coffee have more caffeine?
No. In the Ethiopian dataset covering 850 to 2,775 meters, caffeine showed no correlation with height in either direction, while sucrose, trigonelline and fat all moved strongly. Other studies have found caffeine moving up, moving down, or not moving, which is what a variable with no consistent relationship looks like across a literature.
What does SHB mean on a bag of coffee?
Strictly Hard Bean, a grade given to coffee grown above a stated height in several Central American countries. Published thresholds for Guatemala range from 1,300 to 1,524 meters depending on who you ask. Barista Hustle's curriculum states that quality and size are not part of the Guatemalan classifications, so the letters describe the farm rather than the cup.
Is 1,800 meters in Colombia the same as 1,800 meters in Brazil?
No, because the temperature at a given height depends on how far you are from the equator. Perfect Daily Grind made this point in 2018. A farm near the equator can sit far higher than a farm in a subtropical growing region and still be at a comparable temperature. The meters are a stand-in for the climate, and the stand-in does not travel.
Why do two coffees from similar elevations taste so different?
Because height is one input among several and not the largest. A study of 53 Ethiopian farms found soil chemistry explaining 84 percent of the variation in specialty quality against 3.8 percent for elevation. Variety, processing and drying then move the cup again before it reaches a roaster.
Does slower ripening really concentrate sugar in the bean?
The endpoint is measured and the mechanism is inferred. Green coffee from cooler, higher farms reliably tests higher in sucrose. The slow fill that is supposed to produce it is carried over from earlier plant-science work rather than observed in the coffee studies themselves, so it is a well-supported explanation rather than a measurement.
The coffee this was written about. Kenya Kiambu is the washed Kenyan on our shelf right now, grown in a county our own region page puts between 1,600 and 1,900 meters, and it carries a published cup score of 91.5. We roast to order every Wednesday in Pensacola and grind it to your brewer at no charge.
