Myths are rarely pure invention. Most of them started as sound advice for a particular situation — a freezer full of damp paper bags, a filter that had not been rinsed, a scoop of beans that had been open for months — and then hardened into rules that no longer fit. So each entry below says what is wrong with the claim, and also what the person repeating it is getting at, because usually there is something.
None of this asks you to change how you like your coffee. If a belief you hold is not here, the second section — on how to check a claim — applies to it just as well.
01
Why these claims survive
Coffee knowledge is passed from person to person, with no exam at the end. Most of what circulates is right, and the wrong parts survive because they sound plausible, sell something, or describe something you can genuinely taste.
Start with the mechanism. A good myth usually has one, and it is usually half a mechanism. Caffeine is bitter, dark roast is bitter, therefore dark roast has more caffeine — each step is nearly true and the conclusion is not. Espresso tastes intense, therefore it must be strong in caffeine — true of concentration, false of the cup. Once you have a plausible story, the claim feels like knowledge, and it gets repeated.
Then there is the sales angle, which is rarely cynical but almost always present. “Use it within days of opening” sells coffee. “Dark roast, nice and strong” sells dark roast. A photograph of a thick head of crema sells machines. A vague worry about “chemicals” in decaf sells the caffeinated kind. None of these claims needs to be true to do its job, which is exactly why nobody checks it.
Some of the myths are simply out of date. Decaffeination really did once rely on solvents used carelessly, coffee really was once sold in paper bags and ruined in freezers, and paper filters really were thicker and more flavour-hungry than the ones sold now. The world moved on; the advice did not.
And specialty coffee has its own folklore, which is worth being sceptical about too. Fresh is always better — not if the beans are a day old. Light roast is always better — not if it is underdeveloped. Crema is a fault — not if you like it. The test in every case is the same: what is the mechanism, and what does the cup actually taste like?
02
How to check a claim
Four questions will see off most bad coffee advice, and one of them — the question of units — settles nearly every argument about caffeine.
Ask what is being compared, and per what. Per gram, per millilitre, per cup, per serving, per day? Most caffeine disagreements are unit disagreements in disguise: espresso is far stronger than filter coffee per millilitre and often weaker per drink. Once both people say which they mean, they usually agree.
Ask what the mechanism is. If nobody can say why a thing would be true, treat the claim as a report of preference rather than a fact. “Paper filters strip the goodness” has a mechanism — paper holds back oils and particles — and the interesting question becomes whether those particular oils are good for you. “Beans go off five days after opening” has no mechanism that fits what staling actually is.
Ask who benefits from you believing it, and in what way. This is not an accusation. A roaster who tells you to buy little and often is giving good advice that also happens to be good for them. A supermarket that prints a best-before date two years out is not lying, but it is not telling you anything useful either.
Then test it yourself, because coffee is unusually cheap to test. Buy the same beans twice, change one thing, and taste them side by side: frozen against cupboard, paper against metal, dark against light. Write down what you tasted rather than what you expected. Three or four of these comparisons will teach you more than any page, including this one.
One last distinction, and it is the one people skip: wrong and unimportant are different. The dark-roast caffeine myth is wrong, but the difference it is arguing about is tiny, so it does not matter much either way. The freezing myth is also wrong, and it does matter, because it makes people throw away coffee they could have kept.
03
The claims
Nine things people say about coffee, each with the honest answer and the practical version. Verdicts are blunt on purpose; the paragraphs underneath are where the hedging lives, because several of these are half true in a way that matters.
The claim
“Dark roast has more caffeine — that is why it tastes stronger.”
VerdictWrong, and slightly backwards.
Roasting barely touches caffeine. Dark beans taste stronger because they taste roasted, and a scoop of them actually holds less coffee.
Caffeine is already in the green bean, and it survives roasting more or less intact: heat does not create it, and only very high roast temperatures destroy much of it. What the roast changes is everything around it — the bean loses 12 to 20 per cent of its weight, mostly water, and becomes less dense as it swells.
Measured by weight, the caffeine in a light roast and a dark roast of the same coffee is close, and published comparisons have found differences going both ways: some dark samples slightly higher, because the same caffeine is packed into less mass, some slightly lower at the very dark end. Either way the gap is smaller than the gap between two different coffees, and far smaller than the gap made by how much you use.
Measured by the scoop it is clearer, and it runs the other way. Dark beans are larger and less dense, so a spoon holds fewer grams of them. If you dose by spoon rather than by weight — as most people do — a scoop of dark roast gives you less coffee, and therefore less caffeine, than the same scoop of light.
What makes dark roast taste stronger is roast flavour: smoke, toast, tar and a bitterness that the tongue reads as intensity. Bitterness is not caffeine, and intensity is not strength. The things that actually move caffeine are the species of bean (robusta carries roughly twice as much as arabica), the dose in grams, the brew ratio and the size of the cup.
What to do instead
If you want less caffeine, use less coffee by weight or drink a smaller cup. Choosing dark roast does not lower it, and choosing light roast does not raise it — though a spoon of dark will give you slightly less than a spoon of light.
The claim
“A proper espresso has a thick layer of crema. If it hasn't, something is wrong.”
VerdictWrong. Crema is gas, not a grade.
Crema is carbon dioxide from the roast, held as foam by pressure. It tells you the coffee has gas in it and the machine can build pressure. It tells you nothing about whether the shot tastes good.
Roasting fills the bean with carbon dioxide. Under nine bar of pressure that gas comes out of solution as the liquid leaves the basket, and the oils in the coffee stabilise it into a foam. That is the whole of it. Fresher beans have more gas and give more crema; beans roasted a day ago can be so gassy that the shot pours unstable and frothy.
Some coffees simply foam more. Darker roasts are more porous and hold more gas, robusta produces a denser head than arabica, and certain processes seem to help. So a thin-crema washed light-roast Ethiopian can be a lovely, sweet, complex shot, while a stale robusta blend can produce a magnificent-looking head that tastes of cardboard and burnt rubber. The foam and the quality are almost unrelated.
Crema is also not especially nice to drink. It is bitter, slightly ashy, and it carries a lot of the roast character. Plenty of baristas stir it into the shot or skim it off before serving, and in the cup it settles into a brown ring on the glass. It looks like the point of espresso; it is closer to a by-product.
Two useful things it does tell you. If a shot has no crema at all, the beans are probably old or the machine is not building pressure — both worth knowing. And crema fading within a minute or two is normal, not a fault, because the foam is unstable by nature.
What to do instead
Judge a shot by how it tastes and whether it was brewed to a sensible ratio. If you are choosing between two cafés, ask when the beans were roasted, not how thick the crema is.
The claim
“Once you open a bag, the beans are only good for a few days.”
VerdictHalf true — right instinct, wrong clock.
The clock starts at the roast date, not at the seal. A bag opened at four weeks is usually better than one opened at two weeks and left loose ever since.
Coffee goes stale through oxygen, moisture, heat and light. A sealed bag with a one-way valve is a good place to be: carbon dioxide can escape, oxygen cannot get in. Opening it starts the real exposure, which is why the instinct behind the claim is sound. But a few days is far too short for whole beans kept in a sealed container.
The useful window is measured from the roast. Whole beans are typically at their best somewhere between about one and four weeks after roasting for filter, and often a little earlier for espresso, which benefits from a few days of degassing first and then fades sooner. After the peak they decline gradually — flatter, more papery, less sweet — rather than falling off a cliff on a particular morning.
Very fresh has its own problems. Beans a day or two off the roast are still releasing gas, which makes them brew unevenly and taste sharp, especially under pressure in an espresso machine. A bag roasted yesterday is not automatically better than one roasted a fortnight ago; it is just newer.
Ground coffee is the version of the claim that is true. Ground coffee has vastly more surface exposed to oxygen and goes flat in days, and noticeably within hours by a strict standard. That is the argument for buying whole beans and grinding per brew, not for treating an open bag of beans as a five-day countdown.
What to do instead
Read the roast date, not the date you opened the bag. Buy roughly what you will drink in two to four weeks, keep whole beans sealed, cool, dark and out of the fridge, and grind as you go.
The claim
“Never put coffee in the freezer — it ruins it.”
VerdictOut of date, if you do it properly.
Freezing slows staling and works well for whole beans in an airtight, air-free container. The old advice is about damp paper bags and repeated thawing, which really do ruin coffee.
Staling is a set of chemical reactions, and cold slows chemical reactions down. A domestic freezer at around -18 °C effectively pauses them, which is why roasters, competitors and a lot of ordinary enthusiasts keep coffee frozen. Some coffee is even shipped frozen.
The bad reputation is deserved by a different practice: supermarket ground coffee in a paper bag, put in the freezer, taken out every morning, thawed, refrozen, and quietly absorbing the smell of everything else in there. Condensation on every thaw, moisture in the grounds, odours in the oils — that is a moisture-and-odour problem, not a freezing problem.
The method that works is unglamorous. Freeze whole beans, portioned into airtight bags or containers with as much air removed as you can manage — vacuum if you have it. Take out one portion at a time, leave the rest frozen, and grind straight from frozen: cold beans shatter more evenly, and there is some evidence that grinding them frozen gives a more consistent particle size. Do not refreeze what you have thawed.
Two caveats. Ground coffee freezes less well than whole beans, because there is so much more surface, though it is still better frozen than left in a warm cupboard. And a frost-free freezer cycles its temperature, so the container matters more than the setting. None of this is worth doing for coffee you will drink this week — the point is storing surplus.
What to do instead
Freeze surplus whole beans in airtight portions, grind from frozen, and don't refreeze. What ruins coffee in a freezer is moisture and other people's food, not the cold.
The claim
“Espresso has less caffeine than filter coffee — it's only a small cup.”
VerdictTrue per serving, false per millilitre.
A single shot usually carries less caffeine than a mug of filter. A double, or a large milk drink, often carries as much or more. Per millilitre, espresso is by far the stronger of the two.
The confusion is units, and almost every argument about this is really an argument about units. Per 100 ml, espresso carries several times the caffeine of filter coffee, because it is a concentrate. Per drink, a 30 ml single shot typically lands somewhere around 60 to 80 mg, while a 250 ml mug of filter usually carries roughly 90 to 140 mg. So the small cup usually does have less — not because espresso is weaker, but because there is so much less of it.
The dose decides everything else. A modern double uses 18 to 20 grams of coffee; a filter mug might use 15 to 18 grams in 250 to 300 ml. Same coffee, same caffeine, different amounts of water. That is why a double shot, or a large latte built on one, can comfortably pass a mug of filter — and why a triple-shot iced drink is in a different category altogether.
Species matters more than roast. Robusta carries roughly twice the caffeine of arabica by weight, so a robusta-heavy Italian blend delivers more per gram than a light-roast single origin. Very few people know what is in the blend they are drinking, which is one reason the numbers in the cup vary so much.
There is a fair amount of measurement noise here, and it is worth saying so: published figures for the caffeine in a shot vary by a factor of two between studies, because dose, species, roast, grind and extraction all move it. Treat any single number — including the ones above — as a rough band rather than a specification.
What to do instead
Compare drinks by how many grams of coffee went into them, not by cup size. A single shot is usually gentler than a mug of filter; a double or a large milk drink often is not.
The claim
“Light roast is just under-roasted. It's basically raw coffee.”
VerdictSometimes true of a particular bag; wrong as a definition.
Light roast means the roast stopped early, on purpose. Underdeveloped is a fault, and it can happen at any roast level — including dark.
Two different things are being run together. Roast level is where the roast ended: light means it was dropped just after first crack, before second. Development is how it got there: whether the inside of the bean caught up with the outside. A light roast is cooked through by design, just not taken dark. A dark roast can still be underdeveloped — scorched outside, raw in the middle.
Underdevelopment has a recognisable taste: sour in a way that does not improve with hotter water or a finer grind, grassy or hay-like, sometimes a raw peanut edge, with a drying finish and no sweetness behind it. A well-developed light roast is the opposite — bright, sweet, floral, fruity, with a clean finish. The two are not the same drink, and people who have only met the first one reasonably conclude they dislike light coffee.
The kernel of truth is real, though. The fashion for very light roasts has produced genuinely underdeveloped coffee sold as “bright”, and a buyer who tried one of those and decided light roast means raw is not imagining it. They met a fault and mistook it for a category. The same mistake happens in reverse when a badly baked dark roast is taken as proof that dark coffee is dull.
There is a second reason light roasts get blamed. They are less forgiving: denser, harder to extract, and much less tolerant of a cheap grinder and an unmeasured kettle. Brewed cool or coarse, a perfectly good light roast tastes thin and sour, and the roast gets the blame for the brewing.
What to do instead
Judge a light roast by whether it tastes sweet and fruity or sour and grassy. If it stays sour and papery after hotter water, a finer grind and a longer brew, it is underdeveloped — and that is the roaster's problem, not the roast level's.
The claim
“Decaf is made with harsh chemicals, so it's worse for you than regular coffee.”
VerdictOut of date.
Decaffeination happens on the green bean, before roasting, by water, by carbon dioxide or by a solvent — all regulated, and most modern decaf uses one of the first two. What remains is a trace measured in parts per million.
The process happens before the coffee is roasted. Green beans are decaffeinated, dried, bagged and sold on, and then roasted like any other coffee. Around 97 per cent of the caffeine is removed, which is why decaf still contains a little — a cup typically carries a few milligrams, not zero.
The methods are worth knowing by name, because bags increasingly print them. Water processes — Swiss Water is the best known — soak the green beans and capture the caffeine on an activated carbon filter, with no solvent involved. Carbon dioxide processes use pressurised CO2, which is highly selective for caffeine, and the gas is recovered and reused. Both are common in specialty decaf.
The two solvent methods are where the reputation comes from. Ethyl acetate is a solvent, but the version used commercially is usually derived from sugar cane, and it occurs naturally in fruit and in coffee itself — which is why some bags advertise it as the natural process. Methylene chloride is the one people mean by “harsh chemicals”; it is still used in some large-scale production because it is cheap and gentle on the bean, it is driven off afterwards, and the residue that remains is capped by food-safety regulation at a very low level and typically comes in far below the cap.
The honest framing is that all of these are chemicals, including water and carbon dioxide. What matters is how much is left and whether anyone is checking, and in regulated markets the answer is a trace too small to taste or to matter. What decaffeination does change is the bean: it becomes more porous, roasts differently and goes stale faster, which is why decaf has a reputation for tasting flat and why fresh decaf from a roaster who sells a lot of it tastes far better than the supermarket kind.
What to do instead
Look for the process named on the bag — water, CO2 or sugar-cane ethyl acetate are what most good roasters use — and buy it fresh, because decaf ages faster than regular. If you are avoiding caffeine for a medical reason, remember decaf is not caffeine-free.
The claim
“Coffee dehydrates you, so you need a glass of water with every cup.”
VerdictMostly false, with a small grain of truth.
Caffeine is a mild diuretic, but the water in the cup more than covers it. For most people, coffee counts towards the day's fluid.
Caffeine does increase urine output slightly, particularly at higher doses and in people who do not usually drink it. But a cup of coffee is overwhelmingly water, and the net fluid balance of moderate coffee drinking is close to neutral — the diuretic effect does not outweigh what you just drank.
Habit matters. Regular drinkers develop tolerance to the diuretic effect, and studies of moderate consumption generally find no meaningful difference in hydration markers between people drinking coffee and people drinking water. The research on ordinary intakes is fairly consistent on this, which is unusual for coffee research.
The grain of truth is about edges. Large amounts of strong coffee, in heat, during endurance exercise, or in someone who drinks nothing but coffee and never water, can contribute to fluid loss. Caffeine also irritates some people's bladders and, later in the day, costs sleep — both real effects, neither of them dehydration.
The advice to have a glass of water with each cup is harmless and pleasant, and it is a decent way to notice how much coffee you are actually drinking. It is just not a rescue operation, and treating coffee as a net loss leads people to drink less of it than they would like for no reason.
What to do instead
Drink coffee, drink water when you are thirsty, and don't count millilitres against each other. If you feel jittery or sleep badly, that is the signal to act on — not dehydration.
The claim
“A paper filter strips out the goodness. Real coffee needs a metal filter or a press.”
VerdictIt strips something — mostly the part you would rather not drink.
Paper holds back fine particles and some of the oils. Those oils carry flavour and body, and also the compounds that raise LDL cholesterol. For most people, filtering is the better trade.
Paper removes two things: the finest coffee particles, which is why a paper-filtered cup tastes clean rather than muddy, and a portion of the oils that a metal filter or a French press lets through. Those oils carry aroma compounds and the sensation of body, which is exactly why a French press feels heavier and a pour-over tastes more defined. If you prefer the heavier cup, that is a real preference and not a mistake.
The oils are also where cafestol and kahweol sit — diterpenes that reliably raise LDL cholesterol in controlled trials. Unfiltered methods deliver much more of them: French press, moka pot, Turkish and boiled coffee are the highest, espresso delivers a moderate amount but in a small cup, and paper-filtered coffee removes most of them. This is one of the few places where the compounds being “stripped” are ones you are better off without, and it is why dietary advice on coffee usually points towards filter.
The goodness people worry about mostly survives. Caffeine passes through paper, and so do the chlorogenic acids and most of the polyphenols that coffee is credited with. You are not drinking a depleted cup; you are drinking a clearer one.
What paper does cost you is texture and some aroma, and it can add a papery taste if it is not rinsed — which is the likely origin of the complaint. Rinse the filter with hot water before you brew and most of that goes away. A metal filter keeps the oils and the body but lets the fines through, which is why the last mouthful of a metal-filtered cup can be silty.
What to do instead
Choose paper for a cleaner cup and less of the cholesterol-raising oils, metal or a press for more body and aroma. Neither is stripping the coffee of anything you need.
04
How to use all this
Almost every myth here turned out to be true in a narrow way and false as a rule: crema does indicate freshness, freezing badly does ruin coffee, espresso is stronger per millilitre, paper does remove oils. The mistake is not the observation; it is the leap from a narrow truth to a general law. Once you know which units and which conditions a claim depends on, most of the arguments dissolve.
The short version, if you want one: dose by weight rather than by spoon, check the roast date rather than the opening date, freeze surplus properly if you have surplus, judge coffee by taste rather than by foam or colour, and buy decaf from someone who tells you how they made it. Everything else is a preference you are allowed to have.
Back to the home page — or read Roasts for what roast level really does, Brewing for the extraction logic behind half of these answers, and Espresso at home for crema and pressure properly explained, and Decaf for the decaffeination claims in full, and Coffee and your body for the numbers behind the caffeine claims, with their sources. If a claim sent you looking for a fix rather than an argument, Troubleshooting deals with the bitter, sour and watery cups directly.
