Body is what a cup of coffee feels like rather than what it tastes like: its weight, the film it leaves behind, the drag of the liquid against the roof of your mouth. It is almost never thickness. In 2019 a group at the University of Florence ran eight brewing methods across the same capillary viscometer, with every sample brought to 20 degrees Celsius before it was measured, and the filter brews came out sitting on top of water. Against water at the same temperature, a V60 measured about 1.00, cold brew about 1.01, an AeroPress about 1.02. A French press, the brew with the heaviest reputation of the eight, measured about 0.98 — very slightly thinner than the water it was made with, and thinner than the pour over it is supposed to outweigh. Espresso is the exception and it is a real one. A classic shot came in near 1.23 and a specialty shot near 1.52. Even there, the most concentrated coffee beverage anybody makes tops out around one and a half times the thickness of tap water while feeling like a different category of liquid altogether.
What does body actually mean?
The cupping protocol the Specialty Coffee Association published in 2005 puts it plainly: "The quality of Body is based upon the tactile feeling of the liquid in the mouth, especially as perceived between the tongue and roof of the mouth." Tactile. Not flavor, not aroma, not acidity. Touch.
That definition has been picked over for years, and the best treatment of its vagueness belongs to Jesper Alstrup at CoffeeMind, who quotes the protocol and then asks the obvious question, which is what exactly a taster is supposed to be rating. His own survey of the trade found people meaning at least four different things by the word, with viscosity the most common answer at about 19 percent and an oily coating second.
The research answer arrived in 2023. Brianne Linne and colleagues at Ohio State set out to find what body is actually made of and reported, in the abstract of their paper in Food Research International, that "coffee body was determined to be comprised of 4 sub-attributes including mouthcoating, astringency, chalkiness, and thickness." I should say that this one is paywalled and I read the abstract rather than the paper. The finding is repeated and cited in the group's 2025 follow-up in Food & Function, which I did read in full, so the four sub-attributes are on firm ground even if the 2023 methods section is not something I can vouch for.
Barista Hustle got to the useful version of this before anybody else and credits the Ohio State lab by name while doing it. Their piece lays out the sub-attributes, grounds the whole sensation in touch receptors rather than in taste buds, and makes the argument that matters: these compounds may be acting on receptors directly rather than by making the liquid any thicker. That framing is theirs, and I am not going to restate it as though we arrived at it. What I can add are the numbers.
If coffee is not thicker than water, why does it feel thicker?
It is barely thicker, and in one case not thicker at all.
The Florence measurements come from Angeloni and colleagues, published in Food Research International in 2019, and the reason they are worth more than a dozen assertions is that all eight brews went through one Ostwald-type capillary viscometer at one temperature. Same instrument, same day, same 20 degrees. Expressed against water:
- French press, about 0.98
- V60 pour over, about 1.00
- Cold brew, about 1.01
- AeroPress, about 1.02
- Moka pot, about 1.12
- Classic espresso, about 1.23
- Specialty espresso, about 1.52
I am giving you ratios rather than the paper's printed units on purpose. The units as published do not reconcile against water at the stated temperature, by roughly a factor of a hundred, while the relative ordering is internally consistent and is corroborated in direction by two later papers. The ratios are the trustworthy part, and they are the part that answers the question.
Read the French press line again. The brew everybody reaches for when they want a heavy cup measured as the least viscous of the eight, below a paper-filtered pour over. There is an honest caveat attached: it also had the lowest dissolved solids of the eight, 1.35 percent against the V60's 1.55, so some of that gap is concentration rather than mouthfeel. What the table does prove is the flat fact that filter coffee, whatever you brew it in, is within a percent or two of water. Something else is doing the work.
The reason it still feels heavy is that the things which build body are not thickening agents. They are compounds your mouth detects at concentrations far too low to change how a liquid flows.
The clearest evidence I found is in the 2025 Food & Function paper. The Ohio State group isolated a phenolic-rich melanoidin fraction, all of it under 5 kilodaltons, and found it sensory-active for astringency at the concentrations it actually occurs in brewed coffee: 15.7 milligrams per liter in a high-body coffee and 7.9 in a low-body one. Nine trained panelists could tell it apart from the blank at those levels. Fifteen milligrams in a liter is nothing. It will not thicken water and the paper measured no viscosity at all, because viscosity was never the mechanism under test.
The 2023 work found something stranger still on the mouthcoating side. Two chlorogenic acids produced a perceptible coating sensation, and the perception ran inversely to concentration: more of the compound, less of the feeling. Whatever is happening there is happening at the receptor, or in how saliva behaves, and not in the fluid.
Astringency is the sub-attribute we have already written up in detail, including why a drying finish is a touch sensation rather than a taste, and that piece is the place to go for it. Body is the larger category that astringency sits inside.
Does a stronger cup have more body?
Partly, and this is where I want to be careful, because the tempting version of this article says body and strength are unrelated and that claim is checkable and wrong.
Batali, Ristenpart and Guinard, working at the UC Davis Coffee Center and publishing in Scientific Reports in 2020, put it directly: an attribute they called Viscous "would be expected to scale linearly with TDS, as it refers only to the thickness and mouthfeel of the coffee, and a coffee with more dissolved solids should accordingly have a thicker mouthfeel." Their trained panel found exactly that. Perceived thickness rose linearly with dissolved solids and showed no significant relationship with extraction percentage. So dissolved solids set a floor. Brew the same coffee stronger and it will read as heavier.
What dissolved solids do not do is settle the question. The same group's 2022 paper in Foods ran a study in which every single sample was diluted to precisely 2 percent TDS before anybody tasted it. Concentration held flat by design. Perceived Viscous still moved, and it moved hard with roast level: an F statistic of 182.77 for roast, against 5.36 for origin and 0.87 for brew temperature. Same strength in every cup, and the roast alone changed how thick the coffee felt by a margin that made temperature look like noise.
Put the two papers together and you get something more useful than either slogan. Strength sets the floor on body. Roast moves it independently of strength. And the Florence viscometer says that neither of them is working by thickening the liquid. If you want the strength half properly, our piece on grind and strength walks through what dissolved solids actually are and why a thin cup usually traces back to the weight of coffee you started with rather than to the grinder, and that conclusion still stands. This article is about the half of the sensation that survives after you have held the dose steady.
Does a paper filter or a metal filter give you more body?
Metal, and I am going to own something before I explain why.
We sell paper filters and we sell metal ones, and our own product copy has told customers both things. One page credits a paper filter with better body. Another credits a stainless filter with a fuller cup because it lets the oils through. A third says the same of a metal cone. Those sentences cannot all be right, they are all ours, and nobody buying a filter from us could work out which one to believe. The honest answer is the one the diterpene literature has been sitting on for thirty years.
Cafestol and kahweol are coffee oils, and because they matter to cholesterol research somebody has measured them properly in every brew method there is. Orrje and colleagues published a mass-spectrometry survey in Nutrition, Metabolism and Cardiovascular Diseases in 2025. Median cafestol, in milligrams per liter: paper-filtered drip, 11.5. French press, 87. Percolator, 91. Boiled coffee, 939.
The cleanest line in that table is the next one. Take the same boiled coffee and pour it through a cloth filter and cafestol falls from 939 to 28 milligrams per liter. The coffee did not change. The grind did not change. The brew did not change. A piece of fabric removed about 97 percent of the oil. That is the filter, and nothing else, setting the lipid load of a cup.
So paper strips oils and metal passes them, which is the mechanism Jonathan Gagné describes when he says metal filters produce a beverage with more suspended solids and more body. Our metal-filter copy is right and our paper-filter copy is loose, and I would rather say that here than leave three of our own pages arguing with each other. The loose sentence is going to get fixed.
One thing I cannot give you, and it is worth naming rather than papering over: there is no measurement in any paper I could open of how much insoluble material a given filter actually leaves in the cup. Gagné measured the filters themselves — pore sizes, thickness, flow rates. The oil side is quantified to three significant figures. The fines side is not quantified at all.
Why is espresso the exception?
Because espresso is not the same kind of object as a cup of filter coffee. Ernesto Illy and Luciano Navarini characterized it in Food Biophysics in 2011 as three phases at once: dissolved solids at 52.5 grams per liter, an oil-in-water emulsion occupying 0.2 to 0.3 percent of the volume with 90 percent of its droplets under 10 micrometers, and roughly 5 grams per liter of suspended cell-wall fragments 2 to 5 micrometers across. On top of that sits crema, which is carbon dioxide from the roast held in a foam that accounts for at least a tenth of the cup's volume.
That is a genuinely different liquid, and it is the one case where the thickness you feel is partly thickness you could measure. It is also why the equation between body and viscosity took hold in the first place. Anyone who learned the word behind an espresso machine learned it on the one drink where it happens to be nearly true.
How do you taste body on purpose?
Stop trying to rate it as a single quantity and rate the four things it is made of. World Coffee Research's sensory lexicon does this better than any tasting wheel, because it anchors each one to something you can go and buy. Thickness is "the thick feel of the beverage as you press your tongue through it," calibrated with a 5 percent sugar solution at 2.0 and tomato juice at 4.0. Oily is scored against low-fat milk at 3.0 and half-and-half at 6.0. Mouth drying runs against dilute alum.
Brew the same coffee two ways, paper and metal, at the same dose and the same water. Then ask four questions instead of one. Is it thicker as the tongue moves through it. Does it leave a film. Does it dry the sides of the mouth. Does it coat and then fade, or coat and stay. Those are separable, they move independently, and the moment you stop collapsing them into one word you can tell which of them you actually want.
A note on what our own labels will and will not tell you here. Every coffee we sell carries a field that reads either Espresso or Drip, and that is the closest thing on our shelf to a texture instruction — it is a judgment about which brew we think a lot belongs in, and body is a large part of what goes into it. What we do not publish is the reasoning. You get one word and you have to take it on trust, which is a fair thing to hold against us. If you want the tasting language we do publish and what it is actually derived from, the tasting notes piece covers it, and the ratio piece covers the dose side.
The word body survived a century of coffee writing because it describes something real that everybody can feel. What it never described was thickness. It describes a handful of compounds, present in milligrams, landing on nerve endings that evolved to tell you about texture — and the two levers you actually hold are how dark the coffee was roasted and what you strain it through.
If you want to taste the espresso end of this: ASCENT is the blend we build for the basket, and its record reads Espresso rather than Drip. Pulled as a shot it is the one brew where the thickness is not only a sensation. It is roasted to order on Wednesdays in Pensacola and we will grind it to your brewer at no charge, though for this particular experiment buy it whole and split it between a paper filter and a metal one. ASCENT by Inheritance.
