Perfume concentration is the most quoted number in fragrance and the least binding one. No law in the European Union or the United States says what an eau de parfum is, how much perfume concentrate it must contain, or how it has to differ from an eau de toilette. The tiers are inherited trade habit, printed on glass, sold as precision. What follows is the reference: every tier, the real percentage ranges, the overlap the industry does not advertise, the four instruments that actually bind a fragrance, and the physics that decides longevity and strength once the label has stopped talking.
11 min
The Terms Are Not Regulated. By Anyone.
Regulation (EC) No 1223/2009 is the governing text for every cosmetic product sold in the European Union. It defines what a cosmetic product is, fixes who carries legal responsibility for the safety assessment, mandates the ingredient list, and fills its annexes with prohibited and restricted substances. It does not define eau de parfum. It does not define eau de toilette, extrait de parfum, or cologne. Those words appear nowhere in it as concentration categories, because it is a safety instrument and the tiers are a commercial vocabulary.
The United States is no different. The FDA accepts eau de parfum, eau de toilette, cologne or simply perfume as the statement of identity on a fragrance label, and attaches no percentage to any of them. No international standard fixes them either. The tiers are convention, formed across the nineteenth and twentieth centuries and repeated because the trade found them convenient.
The consequence is the part nobody prints on the box. An eau de toilette from one house can hold more perfume concentrate than an eau de parfum from another, and both labels are entirely lawful. The word is a category. The number is a private decision.
The Concentration Table
The ranges below are the ones the industry uses in practice. They are conventions rather than thresholds, and the third column is the one that does the damage.
| Label on the bottle | Typical perfume concentrate | Overlap with the tier above | What it legally guarantees |
|---|---|---|---|
| Extrait de parfum / parfum | 15–30% | — | Nothing |
| Eau de parfum (EdP) | 10–20% | Overlaps extrait from 15% upward | Nothing |
| Eau de toilette (EdT) | 5–15% | Overlaps EdP from 10% upward | Nothing |
| Eau de cologne (EdC) | 2–5% | Overlaps EdT from 5% upward | Nothing |
| Eau fraîche / brume | 1–3% | Overlaps EdC from 2% upward | Nothing |
Read the third column again. Every tier overlaps the one above it. A 16% eau de parfum and a 16% extrait de parfum are the same liquid under two different names. A 12% eau de toilette outranks a 10% eau de parfum on the single axis the label pretends to describe. The tiers are not a ladder with rungs. They are five overlapping smears, and the house chooses which word to print after the formula is already finished.
This page is the reference. The argument about what that overlap means for what you actually wear lives elsewhere. An eau de toilette can outperform an extrait, and regularly does.
Tier by Tier: What Each Name Has Meant
Eau de cologne is the oldest of the five and the only one named after a place. Giovanni Maria Farina, an Italian perfumer working in Cologne, launched his Eau Admirable there in 1709: bergamot and other citrus oils suspended in rectified grape alcohol, a construction nobody had held stable at that scale before. Its modern descendants sit at 2 to 5% and lean on the same volatile architecture — lemon, neroli, lavender, rosemary. Cologne is the one tier where the low number is the intention rather than a compromise.
Farina did not invent a dilution. He invented a category, and a commercial format that outlived every court it was sold into. How an Italian merchant invented modern luxury.
Extrait de parfum sits at the top, conventionally 15 to 30%. Historically it was not the luxury edition of anything; it was the reference form, the way a composition was first built and first sold, and everything lighter was a derivative of it. Extraits carry a higher proportion of low-volatility material — oud, ambergris and castoreum in the older formulas, jasmine absolute and heavy woods in modern ones — which is why a well-built extrait reads as denser rather than louder.
Eau de toilette, 5 to 15%, began as the day-wear derivative. In the first half of the twentieth century houses started offering one composition at two strengths, pairing the extrait with a lighter, brighter version for daytime. The format widened the customer base far more effectively than the extrait ever had, which is the reason it stuck.
Eau de parfum is the newest tier, not the oldest. Parfum de toilette and eau de parfum labels began appearing in the 1970s and became standard through the 1980s, positioned between the radiance of an eau de toilette and the persistence of an extrait. It is now the default strength at which most new fragrances launch, which is a fact about retail habit rather than about chemistry.
Eau fraîche, brume, body mist sit below cologne, typically 1 to 3%, and carry no historical claim at all. They are commercial descriptors for dilute hydroalcoholic or water-based products. There is nothing wrong with them. Just do not read them as a rung in a system, because there is no system.
What the Law Does Regulate
Four instruments genuinely bind a fragrance sold in Europe or the United States. Not one of them contains the word extrait.
| Instrument | What it actually fixes | Defines EdP or EdT? |
|---|---|---|
| Regulation (EC) No 1223/2009 | Safety assessment, responsible person, ingredient list, banned and restricted substances | No |
| Commission Regulation (EU) 2023/1545 | 80 fragrance allergens declarable above 0.001% in leave-on products | No |
| IFRA Standards, 51st Amendment | Maximum use levels per material, by product category, in the finished product | No — one category covers them all |
| 27 CFR § 20.122 (United States) | Minimum 2 fl oz of perfume material per gallon under denatured alcohol formula 39-C | No — it is a tax rule |
Commission Regulation (EU) 2023/1545, adopted on 26 July 2023, is the one that changed what you read on the back of a bottle. It added 56 fragrance allergens to Annex III of the cosmetics regulation as entries 327 to 371, joining the 24 already listed and taking the declarable total to 80. Any of them present above 0.001% in a leave-on product — and a fragrance is a leave-on product — has to be named individually. Non-compliant products can no longer be placed on the Union market after 31 July 2026, and cannot be made available at all after 31 July 2028. That is why small type at the base of the carton now names eugenol, citronellol, geraniol and coumarin.
The IFRA Standards are the industry's own restriction system, and the 51st Amendment was notified on 30 June 2023. Two details matter here. IFRA sorts products by how they are used, and every alcoholic fine fragrance falls into a single bucket, Category 4, which covers eau de toilette, eau de parfum, parfum and cologne without distinction. And where a Standard imposes a quantitative limit, IFRA states that it is expressed "as an upper concentration of fragrance material in the finished consumer product, not the fragrance mixture."
Follow that second point through and it inverts the usual intuition. The cap applies to the bottle, not to the perfumer's concentrate. A restricted material sitting at 0.4% of a concentrate reaches 0.032% of an 8% eau de toilette and 0.10% of a 25% extrait — the same formula, three times the level in the finished product. Where a restricted material is load-bearing, the higher-concentration version is the one that has to be rebalanced. Oakmoss is the case most people have noticed: restricted, reformulated, conspicuously thinner in the chypres that once leaned on it alongside patchouli, vetiver and sandalwood.
The strangest of the four is American and has nothing to do with safety. Under 27 CFR § 20.122, an article made with Specially Denatured Alcohol Formula 39-C must contain not less than 2 fluid ounces of perfume material per gallon of finished product. Two ounces in a hundred and twenty-eight is 1.6% by volume. That is the only concentration floor written into law anywhere in this article, and it exists so a bottle of cologne is not taxed as a beverage. The companion section, 27 CFR § 21.141, lists perfumes, colognes and toilet waters among the authorised uses for that alcohol.
The law counts allergens and it counts excise. It does not count luxury.
If the tier guarantees nothing, the question becomes what the tier is standing in for. EdT, EdP, parfum: what you are actually paying for.
Why Percentage Does Not Predict Longevity
Longevity is a property of molecules, not of percentages. The governing variable is vapour pressure — how hard a molecule pushes to leave the liquid and enter the air. Limonene, the dominant constituent of cold-pressed citrus oil, is recorded at 200 Pa at 25 °C in its ECHA registration dossier. Galaxolide, the polycyclic musk sitting underneath a very large share of modern fragrance, is recorded at 1.5 Pa at the same temperature in its own dossier. That is a factor of roughly 130 between two materials that can share a bottle at an identical dose.
Run the consequence. A 20% extrait built mostly on limonene, linalool and citral will be off your forearm inside an hour, because 20% of a fast molecule is still a fast molecule. A 6% eau de toilette built on galaxolide, Iso E Super, ambroxan and exaltolide will still be legible at hour eight, because low vapour pressure does not care what the label says. Concentration sets the starting height. Volatility sets the slope.
Molecular weight tracks the same logic. Almost every known odorant sits below roughly 300 daltons, and above that ceiling molecules stop reaching the receptor at all. Emily J. Mayhew and colleagues, writing in PNAS in 2022, assembled a set of more than 1,900 molecules classified as odorous or odorless and found that transport features — volatility and hydrophobicity, the properties that govern whether a molecule can cross from air into the olfactory mucus — predicted odorousness on their own. Reaching the receptor is the bottleneck. Not how much you sprayed.
Projection obeys air movement, not percentage. Warm air rises off skin and carries the light molecules out first. The fluid dynamics of an invisible trail.
Why Percentage Does Not Predict Strength
Perceived strength collapses the percentage argument even faster, and it does so for two independent reasons.
The first is that detection thresholds run across orders of magnitude. Beta-damascenone, the rose-and-stewed-apple molecule that turns up in trace amounts across an enormous range of compositions, carries a reported detection threshold of 0.002 µg/L in water and 0.05 µg/L in a 10 to 12% hydroalcoholic solution, figures attributed to Kotseridis and colleagues. That is parts per billion. Other materials have to be dosed at percent level of the concentrate before they register at all; hedione is routinely used in that range. A formula can be 25% concentrate and still be mostly made of things you barely perceive.
The second reason is psychophysics. Perceived odour intensity follows Stevens' power law: intensity equals a constant times concentration raised to an exponent, and for odorants that exponent sits roughly between 0.2 and 0.8. It is compressive. At an exponent of 0.3, doubling the concentration of a material raises its perceived intensity by about a quarter. Doubling the concentrate in a bottle does not double anything you can smell. It changes the raw material load far more than it changes the experience of wearing it.
At Première Peau we publish one number and hold it across all seven compositions: 20%, extrait strength, every bottle, no tiered ladder. That is a formulation decision rather than a merchandising one — at 20% a composition can carry a heavy natural load without going flat in the second hour. In Insuline Safrine, built around saffron, the concentration is what lets the material stay legible for hours instead of flashing off with the alcohol. One number, seven times, composed and filled in France.
How to Read a Label That Means Nothing
So what does the word on the bottle tell you? It tells you the house's intention about wear context: cologne for the first hour of a hot day, extrait for proximity. It tells you nothing verifiable about how long the fragrance lasts, how far it travels, or how much material is in it. That is not cynicism. That is the legal position set out above.
Three habits make the label usable anyway. First, compare within a house and never across. Inside a single house, the eau de toilette and the eau de parfum of one composition are genuinely two dilutions of a known concentrate, and the tier carries real information. Across houses it is noise.
Second, read the ingredient list rather than the tier. The allergen declarations under Regulation (EU) 2023/1545 are the only compulsory chemistry on the box, and a long allergen list on a fragrance usually signals a heavy load of naturals, because those substances arrive inside the raw materials rather than being dosed neat.
Third, test on skin across a full day. Your skin temperature, the surface you sprayed, and how much you applied will settle the question the label cannot answer.
Dose is the variable you control, and almost everyone gets it wrong in the same direction. Two sprays, and where they go.
The tier system survives because it is convenient shorthand and because nobody has an incentive to dismantle it. Treat it as what it is: a rough statement of intent, printed in French, binding on no one. If you want to know what a concentration actually does, put seven extraits at the same 20% on your own forearm and watch which one is still speaking at hour six. A Discovery Set settles it in the only laboratory that has ever mattered, which is your skin across one long day.
Frequently Asked Questions
What are the perfume concentration levels, from strongest to weakest?
Extrait de parfum at roughly 15 to 30% perfume concentrate, eau de parfum at 10 to 20%, eau de toilette at 5 to 15%, eau de cologne at 2 to 5%, and eau fraîche or body mist at 1 to 3%. Every one of those ranges overlaps the range beneath it. The order is conventional, not legally defined, and two houses can label the same percentage differently.
Is eau de parfum legally defined anywhere?
No. Regulation (EC) No 1223/2009 governs cosmetics in the European Union and never defines the term, and the FDA accepts it as a statement of identity without attaching any percentage to it. No international standard fixes it either. The tier is a commercial convention that each house applies to its own formulas.
Does a higher concentration mean a longer-lasting perfume?
Not reliably. Longevity is governed by the vapour pressure of the individual materials: limonene is recorded at 200 Pa at 25 °C and galaxolide at 1.5 Pa, a difference of roughly 130 times at an identical dose. A 20% extrait built on volatile citrus will fade faster than a 6% eau de toilette built on musks and woody ambers.
What is the difference between extrait de parfum and eau de parfum?
Conventionally the extrait carries more perfume concentrate and a higher proportion of low-volatility material, which makes it read as denser and closer to the skin. In practice the two ranges overlap between 15 and 20%, so a given eau de parfum can be identical in concentration to a given extrait. The reliable difference is the composition, not the word.
What percentage is an eau de cologne?
Modern eaux de cologne generally sit at 2 to 5% perfume concentrate. The format descends from Giovanni Maria Farina's Eau Admirable, launched in Cologne in 1709 and built on bergamot and other citrus oils in rectified alcohol. The low percentage is the intended character of the genre rather than a shortfall.
Why do two fragrances at the same concentration differ so much in strength?
Because detection thresholds vary by orders of magnitude between materials. Beta-damascenone has a reported threshold of 0.002 µg/L in water, while other materials need percent-level dosing before they register at all. Perceived intensity also follows a compressive power law, so doubling a concentration typically raises perceived strength by well under half.
Is there any minimum perfume concentration required by law?
Only one, and it is a tax rule. Under 27 CFR § 20.122 in the United States, an article made with Specially Denatured Alcohol Formula 39-C must contain at least 2 fluid ounces of perfume material per gallon of finished product, which works out at 1.6% by volume. It exists so the product is not treated as a beverage, and it says nothing about quality.
Should I choose the eau de parfum or the eau de toilette of the same fragrance?
Within one house these are two genuine dilutions of the same concentrate, so the comparison is meaningful: the eau de toilette will usually open brighter and settle sooner, the eau de parfum will sit closer for longer. The two are also frequently rebalanced rather than simply diluted, which means the opening can differ. Test both on skin for a full day before deciding.
