Espresso grind size: what the numbers actually tell you

Three reasons, none of them theoretical.

The zero point — where the burrs touch — is not identical between two units of the same model. It is set at the factory within a tolerance, and it moves: new burrs season over the first few kilos, which shifts the whole scale finer without you touching the collar.

Burr geometry changes particle distribution at the same gap. Conical and flat burrs do not produce the same mix of boulders and fines at the same opening, so two grinders set to the same gap do not give the same flow.

And the coffee itself moves the setting. A light roast is harder and denser than a dark one: at the same setting it grinds coarser and runs faster. A bag three days off roast does not behave like the same bag two weeks later.

A click number is a personal reference point — true for your grinder, your coffee, this week. Write it down. Do not lend it.

If you think in fluid ounces, espresso is the one drink where the habit actively hurts you. A double is often quoted as roughly two ounces, and the SCA definition does give a volume: 25 to 35 ml from 7 to 9 g of coffee (14 to 18 g for a double), with clean water at 90.5–96.1 °C forced through at 9–10 atmospheres, "and where the grind of the coffee is such that the brew time is 20–30 seconds."

Read that last clause again. The reference definition of the trade gives no grind value at all. It defines grind by its result: the right grind is whatever produces that time, at that dose, on that machine.

The volume is the problem. Those 25 to 35 ml include crema, and crema is foam — its volume tracks how fresh the coffee is far more closely than how well the shot extracted. Two shots that taste completely different can fill the same line on a shot glass. Weighing the drink in grams removes that variable entirely. And the 7 to 9 g describes baskets smaller than what ships today: 58 mm portafilters are commonly supplied with 18 g baskets, and the 54 mm baskets on many home machines sold in the US and UK sit lower again.

TIMEMORE Black Mirror Basic 2 - Balance à café avec minuteur

A scale that reads to a tenth of a gram and starts a timer on its own changes the whole exercise. Without one you are comparing shots that share neither dose nor yield, and blaming the grinder for a difference that came from scooping.

Particle size is measurable, in micrometres. The measurement is not the problem — the problem is that your collar is not marked in them.

No standard fixes a micron figure for espresso. The 200 to 400 µm range you see quoted everywhere is a trade convention, not a normalised value, and as you just read, the SCA definition does not mention microns at all.

A few manufacturers do publish what one click is worth, which turns dialling in from guesswork into arithmetic:

1Zpresso J-Ultra Moulin à Café Manuel Gris Fer, Fraise Conique

KINGrinder K6

That is the exception, not the rule. Most grinders sold for espresso advertise "40 positions" or "25 levels" and stop there: you know how many clicks you have, not what one is worth. Hence the advice that follows, which needs no manufacturer figure at all — one click, taste, decide.

Here is the arithmetic, and it explains more than any analogy.

A click is an absolute step: the collar moves the burrs a fixed distance, say 16 µm, whatever the starting gap. But espresso lives right at the bottom of the range. If your espresso setting sits around 250 µm, those 16 µm are 6.4 % of the particle size. The same click at a filter setting near 800 µm is only 2 %.

The resistance of a packed bed to fluid flow follows the Kozeny-Carman relation, in which that resistance varies as the inverse square of particle diameter. Applied to both cases, as an order of magnitude:

On espresso a single click moves the shot four seconds and pushes you out the top of a window that is ten seconds wide. On filter the same click vanishes into the normal tolerance of a brew.

Two caveats. This is an order of magnitude, not a prediction: a real coffee bed departs from the model, because fines migrate downward during extraction and because a single channel through the puck short-circuits the whole calculation. And the practical consequence is a concrete buying argument. If you want to move a shot by about five seconds, you are looking for roughly 5 % of particle size — around a dozen microns at 250 µm. A grinder whose click is worth 80 µm cannot aim at twelve microns. It steps over the window, and you go from gushing to choking with nothing in between.

LELIT Fred PL043MMI

This is the case for stepless grinders: you can move by as little as you like. The trade-off is real and rarely stated — with no index, you have no reliable way back to yesterday's setting. A click is a constraint, but it is also a memory.

Pull a reference shot first: fixed weighed dose, 1:2 ratio, timer started with the pump, drink weighed in grams. Then read the table.

That last row wastes more time than any other. A badly distributed bed opens a preferential path, water rushes through it, and the shot comes out short and sour — exactly the symptoms of too coarse a grind. So you correct the wrong way, go finer, make the channelling worse, and go round in circles for a whole bag.

Normcore porte-filtre sans fond 58 mm

A bottomless portafilter settles that diagnosis in one shot, because it makes the underside of the puck visible: a healthy extraction descends as an even cone, while channelling sprays off at an angle and you see it immediately. It is the most direct diagnostic tool on this list, ahead of any prep accessory.

Subminimal WDT Tool

Normcore V4.1 tamper 58,5 mm

For prep, the order is simple: break up clumps in the basket with a needle tool, then tamp level. A calibrated spring tamper does not improve flavour by magic — it makes pressure repeatable between shots, which is exactly what you need when you are trying to isolate the effect of one grind click.

1. Lock the dose. Weigh it to a tenth of a gram and leave it alone until the grind is stable. A dose wandering by 0.5 g is enough to hide a click.
2. Lock the ratio. 1:2 by weight is the most common starting convention: 18 g in, 36 g in the cup. Chase the weight, not a line on the glass.
3. Change the grind by one click, and nothing else.
4. Purge. If your grinder retains coffee between the burrs, the first shot after a change still contains the old grind. Throw a small dose through, or treat the first attempt as void.
5. Taste, log, repeat. Write down the click number, the coffee, the date. That notebook is the only "setting 12" that ever meant anything: yours.

• Degassing. Very fresh coffee releases CO₂ during extraction, which disturbs flow. Behaviour settles over the days after opening, and the setting moves with it.

• Roast level. Lighter, denser, harder to grind: it generally wants finer than the same dose in a dark roast.

• Basket size. Moving from a 14 g to an 18 g basket changes bed depth, and therefore resistance, at an identical grind.

• Machine temperature. An under-heated machine under-extracts, and the sourness that results is indistinguishable from too coarse a grind.

• Burr wear. Slow but real: your setting drifts over months, not over a week.

If you are starting from a bean-to-cup machine rather than a portafilter, the vocabulary changes but the reasoning does not — our bean-to-cup comparison and beginner's guide go into what the built-in grinder is actually worth model by model.

TIMEMORE Chestnut C3s PRO - Moulin à café manuel

One last note if you are starting out: a manual conical-burr grinder is more than enough to learn dialling in, because it gives you clean clicks and a usable range. What you buy by moving up is not "better coffee" as such — it is a finer step and a repeatability that wastes less coffee while you are searching.