Learn · Guides
What Is Crema, Really?
Crema is the reddish-brown foam that sits on a fresh espresso shot: carbon dioxide from the roast, whipped into a fine emulsion with the coffee's oils. It looks like a quality badge, but it is one of the most misread signals in coffee.
We earn a commission when you buy through our Amazon links, at no extra cost to you. It never changes a verdict, and we say so when the cheaper product is the better buy. How this works.
Crema is the layer of reddish-brown foam that forms on top of a fresh espresso shot. It is not a separate ingredient and it is not added — it is made on the spot, in the few seconds the shot is pulling, out of things already in the coffee. Under roughly nine bars of pressure the water forces carbon dioxidegas that was trapped in the beans during roasting out of solution, and that gas whips up with the coffee's oils and dissolved solids into a fine, bubbly emulsion. When the pressurised liquid leaves the basket and hits normal air, those tiny gas bubbles are held in place by the oils and proteins, and they float to the surface as the familiar tiger- striped cap. That is all crema is: espresso's own dissolved gas and oils, foamed by pressure. Learn to read it and it tells you a little; treat it as a grade and it will mislead you constantly.
Here is the part most people get backwards. Because good espresso usually has crema, it is tempting to run the logic in reverse and assume that more crema means betterespresso. It does not, and the reasons are not subtle. The single biggest driver of thick crema is not skill or quality — it is how much CO2 the coffee is carrying, which depends mostly on the bean species, how dark it was roasted, and how fresh it is. A cheap, dark, robusta-heavy blend can throw a huge, durable head of crema while tasting harsh, and a superb, delicate, lightly roasted single origin can pull with only a thin, pale layer. If you judged those two shots by their foam you would rank them exactly upside down. The look of the cup and the taste of the cup are only loosely connected, and the whole point of this guide is to keep the two separate in your head. When something tastes off, chase the flavour — our guide to why espresso tastes sour or bitter is a better troubleshooter than staring at the crema.
None of this means crema is useless — it is just a weak, easily faked signal that gets treated like a strong, reliable one. It can hint at freshness, it changes colour with extraction, and its absence sometimes points at a real problem. But it can also be manufactured on demand: a pressurised (dual-wall) basket aerates the coffee through a tiny hole and produces a thick, foamy crema from almost any grounds, stale supermarket coffee included, which is exactly why beginner machines ship with them. Below we cover what crema actually is, what it can and cannot tell you, how grind, freshness and roast move it around, and why the honest way to judge a shot is to pull it properly and then taste it.
What crema actually is (the science, briefly)
Roasting does two things that matter here. It develops oils and aromatic compounds, and it generates a lot of carbon dioxideinside the bean, most of which stays trapped in the bean's structure until you grind and brew. When you pull a shot, the pump pushes water through the packed puck at around nine barsof pressure — roughly nine times atmospheric pressure. At that pressure the water is supersaturated with the coffee's dissolved gas and it strips oils out of the grounds. The instant that pressurised, gas-and-oil-laden liquid exits the basket into open air, the pressure drops, the dissolved CO2 comes rushing out of solution as millions of tiny bubbles, and the coffee's oils, proteins and melanoidins (the brown compounds formed in roasting) coat those bubbles and hold them together. The result is a foam — technically an emulsion of gas in liquid, stabilised by surface-active molecules. That is why crema sits on top, why it is finely bubbled rather than soapy, and why it slowly collapses as the gas escapes: nothing is holding the bubbles forever.
Two practical consequences fall out of that. First, crema needs pressure to form, which is why brewed coffee, pour-over and a moka pot do not make it — they never reach the bar pressure that dissolves and then releases the gas so violently. Second, crema needs gas to form, which is why it is so tightly linked to freshness: coffee sheds its CO2 steadily after roasting, so the same beans that gushed crema in week one will manage only a thin layer a couple of months later, even pulled identically.
What crema can (and cannot) tell you
Read modestly, crema offers a few genuine hints. A shot with essentially nocrema from beans you know are fresh is worth investigating — it can point to very stale coffee, a grind far too coarse, or channelling that let water gush straight through. The colour of the crema shifts with the pour: a shot that runs blonde and pale very early tends to be over-extracted and thin, while a dark, reddish, hazelnut crema usually accompanies a shot pulled in a sensible time. Watching the stream go blonde is a real cue that pros use to end a shot.
What crema cannotreliably tell you is whether the espresso tastes good, because its thickness is dominated by variables that have nothing to do with the cup. It cannot tell you the shot is well balanced, that the grind is dialled in, or that the coffee is high quality — a foamy head sits happily on top of a sour, under-extracted shot or a harsh, bitter one. Treating crema volume as a score is like judging a beer by the size of its head: it is real, it is visible, and it is mostly beside the point. The only instrument that grades a shot is your tongue.
How grind, freshness and roast change your crema
Three inputs move crema far more than your technique does:
- Freshness. This is the big one. Freshly roasted beans are loaded with CO2 and produce abundant crema; as they degas over days and weeks, crema thins out. Beans rested a few days to a couple of weeks after roasting often pull the best-tasting and best-looking shots — very fresh coffee can actually be so gassy it causes uneven extraction. Long-stale supermarket beans give little to no crema on a normal basket.
- Roast level.Darker roasts develop more surface oils and, generally, throw a thicker, darker crema; lighter roasts are denser, less oily and give a thinner, paler layer. That is a difference in style, not quality — a great light roast is supposed to have modest crema.
- Bean species. Robustatraps and holds CO2 far more effectively than arabica and produces noticeably more crema, which is why traditional Italian espresso blends add robusta and why the crema-heavy look became the espresso ideal. Robusta also tends to taste harsher and more bitter, so "more crema" and "better flavour" can literally pull in opposite directions.
- Grind.A finer, correctly dialled grind builds the resistance that lets pressure develop and crema form; too coarse and water rushes through with little crema. Grind is the lever you actually control — see espresso grind size and how to dial in espresso.
The pressurised-basket shortcut (fake crema)
If crema were an honest quality signal, you could not manufacture it — but you can. Most entry-level machines ship with a pressurised, or dual-wall, basket: a second wall with a single tiny hole that the coffee is forced through. That hole creates artificial back-pressure and aerates the coffee jet on the way out, whipping up a thick, bubbly crema from almost anything, including coarsely ground or stale beans that would otherwise give none. It is a helpful training wheel — it makes forgiving shots for people without a good grinder — but the crema it produces is a mechanical foam, not a readout of the coffee. It tends to be big-bubbled and to dissipate faster than the fine, persistent crema from a proper unpressurised basket. The full trade-off is in our guide to pressurised vs unpressurised baskets. The takeaway for judging crema: if you do not know which basket produced a shot, its crema tells you almost nothing.
Why crema is not a reliable quality signal
Put the pieces together and the conclusion is unavoidable. Crema volume is driven mostly by gas content (freshness), roast darkness and bean species, and it can be faked outright by a pressurised basket — none of which track how the shot tastes. You can have gorgeous crema on a mediocre cup and thin crema on a wonderful one. So use crema the way a mechanic uses a warning light: a total absence from fresh beans is worth a look, and a stream going blonde early is a useful cue to stop the shot, but the amountof foam is not a grade. Judge espresso the honest way — dial it in, weigh your dose and yield, watch the time, and above all taste it. If you want to chase real quality, spend the attention on a good grinder and fresh beans, not on chasing a thicker head of foam. Our best coffee beans for espresso and best grinders for espresso pages are where that effort actually pays off, and our methodology explains how we weigh signals like this.
Questions
Frequently asked
What is crema made of?
Does thick crema mean the espresso is good?
Why does my espresso have thin crema or none at all?
Why does cheap or robusta coffee make so much crema?
Do pressurised baskets make fake crema?
Keep reading
Related
- Pressurised vs unpressurised basketsHow dual-wall baskets manufacture thick crema from any grounds — the reason crema can be faked.
- How to pull an espresso shotThe honest way to judge a shot: dose, yield and time, then taste — not the size of the foam.
- Why is my espresso sour or bitter?Chase flavour, not crema — how to read a bad shot by taste and fix the real cause.
- Best coffee beans for espressoFreshness and roast drive crema more than technique — where the effort actually pays off.
- How to dial in espressoGrind and dialling in build the pressure that forms real crema — and the balance that makes it taste right.
Receipts
Sources
- Home-Barista — community reference on crema formation, freshness and extraction
- Specialty Coffee Association — espresso extraction and quality guidance
- Breville — espresso extraction and basket (single vs dual-wall) resources
We do not run a testing lab, and we do not pretend to. Where a measured number came from someone else's work, we name them and link them. Where we could not verify something, we say so on the page rather than quietly leaving it out. Read our full method.