Decaf: how caffeine is removed from coffee and how much is left
Decaffeinated coffee is not coffee without caffeine, but coffee where it makes up no more than 0.3% by dry weight. In a cup, this comes out to between 2 and 15 milligrams — compared to over a hundred in a regular one. Zero does not appear in any document in this story.
The limit is 0.3%, and it is the same for all types of coffee
In Russia, this value is set by product standards. GOST 32775-2014 "Roasted coffee" requires regular roasted coffee to contain at least 0.7% caffeine calculated on a dry matter basis, and sets a limit of no more than 0.3% for decaffeinated coffee. For instant coffee, the lower thresholds are different while the upper threshold remains the same: GOST 32776-2014 requires at least 2.3% caffeine in regular coffee and no more than 0.3% in decaffeinated coffee.
In the European Union, the same 0.3% threshold is specified for instant coffee and coffee extracts: Directive 1999/4/EC permits the term "decaffeinated" if the anhydrous caffeine content does not exceed 0.3% of the dry matter mass of the coffee-based product, and requires this term to be placed in the same field of vision as the product name.
Two clarifications to prevent misattributing rules to these documents. For roasted whole bean and ground coffee, there is no single European figure—Directive 1999/4/EC does not apply to it, and national rules apply instead. In the US, there is no definition of "decaffeinated" at all: there is no standard of identity for coffee in federal regulations, and the common phrasing "97% of caffeine removed" does not correspond to any agency requirement.
Zero in this scheme is unattainable and not required by anyone: decaf is coffee in which the caffeine content is two to eight times lower than the minimum allowed for regular coffee. The standard does not specify the exact amount, it only sets the ceiling.
Three methods, and only one of them is regulated
Caffeine is removed from green coffee beans before roasting. The methods differ in what substance is used to wash it out, and from a regulatory standpoint, this is the key difference: solvents are divided into two categories.
Directive 2009/32/EC on extraction solvents is structured as follows. Its first part lists solvents used "in accordance with good manufacturing practice" without numerical limits: propane, butane, ethyl acetate, ethanol, carbon dioxide, acetone, and nitrous oxide. The condition for them is formulated in words rather than numbers: residues are permitted only in technically unavoidable amounts that do not pose a hazard. Separately, Article 2 of the directive authorizes water as an extraction solvent.
The second part lists solvents with specified residue limits, and methylene chloride is located there.
| Method | Washing agent | EU residue limit |
|---|---|---|
| Solvent | Methylene chloride (dichloromethane) | 2 mg/kg in roasted coffee |
| Solvent | Ethyl acetate | no numerical limit, good manufacturing practice only |
| Water | Water and carbon filter | not applicable, water is not regulated |
| Carbon dioxide | Pressurized CO₂ | no numerical limit, good manufacturing practice only |
Under the same directive, methyl acetate and ethyl methyl ketone are approved for decaffeination with a limit of 20 mg/kg, while the methylene chloride limit for tea is higher than that for coffee at 5 mg/kg.
What can be said about the processes themselves without referencing figures: in the water method, caffeine passes into water and is then trapped on a carbon filter; in the carbon dioxide method, compressed CO₂ plays that role. In both cases, chlorinated solvents are absent from the chain—which is why no residue limit has been established for them. Specific values for pressure, temperature, and the proportion of extracted caffeine are not present in open agency documents, and manufacturers describe them themselves; these numbers cannot be verified against primary sources, which is why they are omitted here.
Manufacturers are not required to indicate the method on the package. If it matters, the only guidance is what the manufacturer stated voluntarily—phrases about water or CO₂ are usually placed on the front panel precisely because they serve as a selling point.
Methylene chloride: how much is permitted and what is known about it
Residue limits in the two jurisdictions differ fivefold, and both figures are small. In the EU, it is 2 mg per kilogram of roasted coffee. In the US, regulation 21 CFR 173.255 permits a residue of no more than 10 parts per million (0.001%) in decaffeinated roasted coffee and soluble coffee extract.
The following is known about the substance itself. The International Agency for Research on Cancer in 2014 moved dichloromethane from Group 2B ("possibly carcinogenic to humans") to Group 2A—"probably carcinogenic to humans", Volume 110 of the monographs.
What this classification means and what it does not mean. The IARC group describes the property of a substance rather than the risk from a specific dose: the evaluation was based primarily on occupational inhalation data. Neither IARC nor EFSA has a separate risk assessment for a residue of 2 mg/kg or 10 ppm in a cup of coffee—EFSA has no opinion on dichloromethane in food whatsoever. Therefore, the honest formulation here is precisely this: there are established residue limits and there is a hazard classification for the substance; neither of these documents contains a conclusion that "coffee with such a residue is harmful" or "safe."
For whom this might be a reason to choose the water method or CO₂: for someone who prefers not to deal with chlorinated solvents in principle. This is a preference-based decision rather than a safety one, and presenting it otherwise would be incorrect.
How much caffeine remains in a cup
Percentages from standards translate poorly to a cup, so measurements are more useful. The US Food and Drug Administration cites the following amounts: decaf contains from 2 to 15 mg per 8 fluid ounce cup (about 237 ml), while regular brewed coffee contains from 113 to 247 mg per 12 ounces.
The nutrient database from the same department provides a comparable picture: 1 mg of caffeine per 100 g of brewed decaf (entry 171889) versus 40 mg for regular brewed coffee (entry 171890)—meaning about 2 mg per cup versus approximately 95.
Two conclusions follow from these numbers, and the second one is frequently overlooked.
First. The difference is fortyfold on average. For a person who eliminates caffeine for the sake of sleep or well-being, decaf solves the problem.
Second. The spread within decaf ranges from 2 to 15 mg, which is a sevenfold difference. Five cups of decaf at the upper limit provide up to 75 mg of caffeine: less than a single regular cup, but no longer "nothing." If caffeine is eliminated for medical reasons and every milligram counts, this is worth considering. How much caffeine is in other beverages and how it generally works is covered in the breakdown on coffee and on adenosine; tea has its own arithmetic.
Decaffeination does not affect calorie content: black coffee has almost no calories anyway, and at a coffee shop, you need to count milk and syrup.
For whom the difference truly matters
During pregnancy. Here there is a figure, and it is specific. The European Food Safety Authority in its scientific opinion on caffeine safety (2015) specified up to 200 mg per day from all sources for pregnant and breastfeeding women—twice lower than the 400 mg for healthy adults. The UK National Health Service cites the same figure and breaks it down by beverage: about 100 mg in a cup of instant coffee, 75 mg in a cup of tea, 40 mg in a can of cola, and 80 mg in a 250 ml can of energy drink. Decaf occupies an almost zero position in this arithmetic—and that is precisely why it is useful here: it allows you to avoid counting. What else changes in the diet during this period is discussed separately.
For sleep issues. The same EFSA opinion contains a precise formulation: a single dose of about 100 mg (approximately 1.4 mg per kilogram of body weight) may affect sleep duration and structure in some adults, especially if consumed close to bedtime. One hundred milligrams is one cup. Decaf at the upper limit of its range does not reach this amount, and an evening cup ceases to be an issue.
Indirect confirmation is provided by the randomized CRAVE trial: among one hundred adults, coffee compared to caffeine abstinence over two weeks reduced nighttime sleep by an average of 36 minutes (397 versus 432 minutes, 95% confidence interval 25–47 minutes).
For arrhythmias. Here expectations and data diverge, and this should be stated plainly. In the same CRAVE trial, coffee did not increase the number of premature atrial contractions—58 versus 53 per day, rate ratio 1.09 with an interval of 0.98–1.20—but it did increase premature ventricular contractions: 154 versus 102, ratio 1.51 with an interval of 1.18–1.94.
Meanwhile, the DECAF trial tested precisely the scenario usually recommended. Two hundred people with persistent atrial fibrillation after cardioversion were divided for six months into those who drank at least a cup of coffee per day and those who completely quit both coffee and decaf. Recurrence occurred in 47% of those drinking coffee versus 64% of those who quit; risk ratio 0.61 with an interval of 0.42–0.89. In other words, quitting coffee in this group did not help, but was instead accompanied by a higher recurrence rate.
The conclusion from this pair of studies is cautious and the only one possible: the assertion that "caffeine causes arrhythmia" is not supported by data, the picture depends on the type of rhythm disorder, and switching to decaf for the sake of the heart showed no benefit in itself. If a rhythm disorder has been diagnosed, this is a question for your cardiologist, not for packaging choices.
What follows from this
- Decaf is up to 0.3% caffeine, not zero. A cup contains 2 to 15 mg versus 113–247 mg in a regular cup.
- Manufacturers are not required to indicate the method on the package. Water and CO₂ are usually voluntarily displayed on the front panel; the absence of such text conveys nothing.
- Solvent residue limits are small: 2 mg/kg in the EU, 10 ppm in the US. No agency holds a risk assessment for these values.
- During pregnancy, the target amount is 200 mg per day in total. Decaf practically frees up this quota.
- For sleep, the guideline is a single dose of 100 mg. Decaf does not reach this level even with five cups.
Frequently asked questions
- How much caffeine remains in decaf?
- According to standards, no more than 0.3% on a dry matter basis: this is specified by GOST 32775-2014 for roasted coffee and GOST 32776-2014 for instant coffee, and the same amount is permitted by EU Directive 1999/4/EC for coffee extracts. In a cup, according to the U.S. Food and Drug Administration, this is 2 to 15 mg per 237 ml, compared to 113–247 mg in regular brewed coffee.
- How is caffeine removed, and is methylene chloride dangerous?
- By four methods: methylene chloride, ethyl acetate, water with an activated carbon filter, and pressurized carbon dioxide. EU Directive 2009/32/EC limits methylene chloride residue to 2 mg/kg in roasted coffee, and U.S. rule 21 CFR 173.255 limits it to 10 parts per million. The substance itself was classified by the International Agency for Research on Cancer in 2014 as Group 2A — probably carcinogenic to humans — but this is a hazard classification of the substance, not a risk assessment of the residue in a cup: neither IARC nor EFSA has such an assessment.
- Can decaf be consumed during pregnancy?
- The restriction during pregnancy applies to caffeine in general: EFSA's 2015 opinion indicates up to 200 mg per day from all sources, half the norm for healthy adults, and the UK National Health Service names the same amount. Decaf provides 2 to 15 mg per cup, meaning it barely falls into this category. Dietary decisions during pregnancy should be made with a doctor.
- Does switching to decaf help with arrhythmia?
- There is no data to support this. In the CRAVE trial, coffee did not increase the number of atrial premature contractions but did increase ventricular ones. In the DECAF trial, among 200 people with persistent atrial fibrillation, complete abstinence from coffee and decaf was associated with a higher recurrence rate than consuming at least one cup per day: 64% versus 47%. For an established heart rhythm disorder, this is a question for a cardiologist.
Read next
- Coffee: how much you can drink and what adds calories — There are three questions about coffee, and they are different: how much caffeine is safe, does coffee harm your health, and why a latte with syrup costs as much as a full meal.
- Caffeine and adenosine: why it works, why it stops, and what to do about it — Caffeine does not supply energy.
- Tea: how much caffeine is in it and what is known about its benefits — Black, green, oolong, and white teas are all the same leaf from the same plant, differing only in the degree of oxidation.
- Nutrition during breastfeeding: how much do you really need — A long list of restrictions has formed around the diet of a nursing mother, which is almost entirely unsupported by evidence, alongside a very short list of what is actually important.
This article is for general information. It is not medical advice, a diagnosis, or a prescription for treatment or a diet, and it does not replace a consultation with your doctor. If you have a health condition, are pregnant, take medication, or follow a diet prescribed to you, decisions about food belong with your doctor.
Figures from regulations, guidelines and studies are given as they stood when this article was prepared and may since have changed; check them against the primary sources. This article is not advertising, an offer, or individual advice, and neither the author nor the site owner is responsible for decisions taken on the basis of it.